A CPU cooler is one of the simplest components in your PC, but it can be a real headache trawling through reviews and opinions as to which options are best for you. To help inform your decision making we obtained a selection of entry-level coolers along with the stock options supplied with some Intel and AMD CPUs, and tried them out.

Most of the entry-level air coolers take the format of a tower of heat fins, which dissipate heat carried to them via heat pipes: These copper pipes contain volatile fluid that boils at the ‘cold plate’ and then re condenses at the heatsink. This mechanism uses the energy absorbed in the phase change of the fluid to absorb and transfer heat. The fan and number of heat pipes are the primary variables that dictate the ability of the cooler, and the fan itself defines how noisy or quiet the cooler is. 

We tested for ease of fitting, cooling capability with an appropriate CPU, noise as represented by fan speed, as well as identifying some compatibility issues with one of the coolers here. By the end of this article, you’ll have all the information you need to choose a cooler for a Ryzen 3600 or 5600X or an Intel i5-10400 to an i5-11600K CPU.

We’ve produced a couple of graphs showing the logged data when running these coolers under an all-core load with an Intel Core i5-11500. This chip draws 110W in this test, enough to expose any Coolers that aren’t up to the task. Sadly that includes the cooler that comes in the box with this CPU. 

If you’re interested in more power hungry CPUs requiring more robust cooling solutions, please check out our article on the best high-end air coolers here. We’ve also produced accompanying videos so take a look over on our youtube channel here.

Stock Coolers: Up to the job?

1. AMD Wraith Stealth

AMD supplies the ‘Wraith Stealth’ CPU cooler with their Ryzen 3600 and 5600X. It’s a simple cooler consisting of an aluminium heat sink, with a shrouded fan blowing down over it. Fitment is simple, only requiring removing the plastic ‘hook’ style retaining brackets, and screwing into the existing Motherboard backplate with the sprung screws. Thermal paste is pre-applied, but if you make a mistake in fitting you may want to re-paste before refitting the cooler.

Cooling performance is only acceptable. It doesn’t throttle a Ryzen 3600 under all-core load, but temperatures are high. Fan speeds are also high, meaning this cooler makes an intrusive noise that you will hear in most instances. It’s not acceptable for the Ryzen 5600X, a higher performance CPU with will benefit from a more capable cooler. We’d recommend checking out the affordable tower coolers later in this test if you’re looking at that CPU. A cost-effective solution is to look for an unwanted Wraith Prism from a Ryzen 7 build: These coolers use heat pipes and are much more effective, able to cope with a 95W TDP CPU without issue.

2. Intel Stock Cooler

The Intel Stock cooler has been updated for Rocketlake era 11^th generation CPUs and is supplied with non-K series i5 options. It now has a black plastic frame and a copper core to assist in heat transfer away from the CPU. Sadly, it appears these changes haven’t been enough. Testing this CPU cooler drove the i5-11500 to its 96C limit and thermal throttling was the result, preventing the CPU from hitting its all-core clock speeds and resulting in lower performance. This was on an open test bench, the situation will only be worse in a closed case no matter how good the airflow. Meanwhile, the fan achieved 3400RPM, generating an irritating buzz. We do not recommend that you use this cooler on any CPU more demanding than an i3, and even in that case you will benefit from a cooler with less intrusive fan noise when it’s working hard. 

This cooler only serves to highlight the inefficiency of Intel’s latest CPUs. It should not be supplied with them, a few dollars reducing in MSPR to put towards an adequate aftermarket cooler would be more welcome.

Best Entry Level Air CPU Coolers

1. BeQuiet! Pure Rock Slim

Moving to aftermarket tower coolers with 92mm fans, the Bequiet! Pure Rock Slim is a cost-effective option with 3 heat pipes and a brushed silver backplate. Fitting is somewhat fiddly on intel, with a pair of brackets that must be screwed into the cold plate, and then plastic push pins to secure it that do not give the most robust installation force. On AMD, fitting is much simpler with the use of a retaining bar that latches to the standard AMD retaining brackets. Thermal paste is pre-applied so do your best to fit it right the first time! Of note is that on an AMD CPU it will only fit with the fan facing up/down, not the more traditional ‘pointing towards the rear’ so if the aesthetics are important that may not be to your liking.

In operation, this cooler proves capable with mid-range CPUs. It maintains 80C under all-core load, and fan speeds stay under 2000RPM and the noise does not become intrusive. It appears that BeQuiet! have prioritised the sound profile over outright cooling performance but this is a balanced cooler that will deal with the Ryzen 3600 or intel i5 non-K CPUs well, whilst keeping costs down. 

Overall this is a cheap (at $20-$25) cooler with good looks and reasonable performance and we have no hesitation recommending it for an entry-level build.

2. Arctic Freezer 7X

Arctic have a good reputation in the cooling segment with great value Fans and high performance all in one water loops, so we were eager to try out the Artic Freezer 7X. This cooler uses a 92mm fan like the Bequiet Pure Rock Slim but has just 2 heat pipes, although they are in direct contact with the CPU heat spreader. The fan is aggressively profiled and ducted, whilst the sides of the radiator fins are enclosed to create a tunnel for air to be forced through.

Our first challenge with this cooler came in fitting it to our Intel test bench, using the ASUS TUF B560M Motherboard. The cooler has two sprung arms to latch onto a plastic fitting ring, and they were too wide and interfered with VRM and M.2 heatsinking in both orientations. There is an offset angle orientation but the cold plate does not cover the whole CPU surface is mounted in that orientation. This was also a problem with the MSI B560M Bazooka, and whilst it is possible to fit the Gigabyte B560M Aorus pro it means not fitting the Primary M.2 heatsink as it interferes with the retaining clip. We could find no note of a compatibility list on the Arctic site, and our email request to support remains unanswered a week later. Instructions aren’t supplied, there’s a QR code directing you to the Arctic site and the instructions don’t make mention how to adjust the fitting ring for different Intel sockets. 

Fitting on AMD is mercifully simple and universal, using the standard AMD retention clips. 

Once fitted, the cooler performed well. It maintained 70C on our test CPU with the fan at just over 2000rpm, and the noise was audible but not intrusive. That it achieves this performance with just 2 heat pipes is to the credit of the design, the fitting mechanism is to its detriment.

Whilst we had no problems with the performance of this cooler, the confusion and incompatibilities over the Intel fitment were a big disappointment, and as such we can’t recommend it for any Intel system: You absolutely want good motherboard VRM heatsinking for the power-hungry intel i5 CPUs, but there’s no guarantee this cooler will fit, and no easy way to check. Arctic need to provide this information to customers to help them make their purchasing decision. 

For AMD, it’s a good, cost-effective option with decent performance, and it will cool a Ryzen 3600, 3700X, or 5600X acceptably well. We can’t recommend it for Intel due to the fitting issues we encountered, so look to the BEquiet! Pure Rock slim as an alternative, or else the 120mm coolers that follow in this test.

3. Cooler Master Hyper 212 Evo V2

This cooler shows it’s age mainly in the hardware it’s supplied with. There’s a large range of posts and brackets to enable it to fit to a number of older intel sockets. This leads to a slightly confusing installation process where the back plate must be configured correctly, and the right stand offs selected for the socket type. The back plate is plastic and feels somewhat flimsy until braced by the cooler itself. The instructions are reasonable, but indicate that you should apply the thermal paste early in the process – a sure fire way to get it spread all over the socket as you fit brackets prior to mounting the cooler itself. 

Performance is good. The fan maintains 1200RPM and noise is low. There’s enough thermal headroom to tune the fan profile to your liking, for either lower noise or lower temperatures at the expense of higher fan speed but at our default test settings temperatures remained under 70C. 

Overall this cooler is still a good performer, but is let down by the legacy fitting options confusing the install process, and by newer designs offering better value and a higher quality package. At the right price, it’s still a solid buy.

4. ID Cooling SE-224-XT RGB

The ID Cooling SE-224-XT is available in a number of configurations, with a basic version, black or white version, and this RGB Version. The cooler is the same basic design throughout with four heat pipes and a 120mm fan compatibility and ID Cooling quote a 180W Capacity.

In the box you get a tube of thermal paste, RGB Splitter to allow the fan and RGB ‘face plate’ to synchronise, and even a stand alone SATA powered RGB Controller if your motherboard doesn’t have 12V RGB compatibility. They also include spare fan clips to enable easy fitment of a second fan.

Fitment is straight forwards, with a solid metal backplate supplied, and a mounting mechanism that works in the same way as Noctuas. There are no RAM or GPU interference issues although Ram with thick heat spreaders will likely touch the fan. Only the RGB on this version slightly complicates set up and install, with a little extra cable management required, but that’s to be expected with any RGB components. The instructions are straightforward to follow.

Performance is good, matching the Hyper 212 with the fan at 1200 RPM and CPU temperatures maintained in the high 60’s. 

Overall this cooler makes a strong case for itself in this segment. It’s attractively priced, with a variety of versions to suit your preferences for color, RGB or cost. The range of accessories is impressive particularly this RGB models inclusion of a stand-alone RGB controller to ensure you can integrate it to any PC build, and we like to see the generous tube of thermal paste included as well: That’s a bonus if you need to refit the cooler for any reason. ID-Cooling has clearly given a lot of thought to the design and implementation of this cooler and it all adds up to a very attractive package, and meets the needs of this section of the market perfectly. It gets the ‘Best in Test’ award for this entry-level cooler roundup.

Fan Speeds and Thermal Performance Testing

Taking a quick look at the data we logged in our testing, we can see the different tiers of performance on offer. This test involved running Cinebench R23 on a 10-minute loop. This CPU pulls 110W in this workload.  We’re showing the first five minutes here as nothing particularly interesting changes after that. The AMD Wraith Stealth is excluded as it is a different workload.

The Intel Stock cooler shows the worse temperatures, with the Intel i5-11500 hitting 96 °C and throttling the CPU. The Bequiet! Pure Rock Slim maintains a much more reasonable 80°C, and does this with no noise. The Arctic Freezer 7X manages to maintain 70°C despite it’s 92mm fan and two heat pipes. 

The two larger coolers are both equipped with four heat pipes and a single 120mm fan. They perform almost identically in this test. Both maintain 65-67°C throughout.

The fan speeds clearly differentiate between the fan size, primarily. The Intel cooler has the smallest fan and spins it the fastest, 3400RPM, whilst failing to adequately cool the CPU. Both 92mm fans spin at around 2000RPM under load, but both remain impressively quiet despite this.

And both 120mm fans maintain 1200 RPM whilst cooling the CPU, meaning they are much less intrusive.

Conclusions and other options

This testing showed the value of a modest upgrade from the stock Intel and AMD CPU coolers. The AMD Wraith Stealth is marginal, whilst the Intel Stock cooler is simply unacceptable for the intended CPUs. On all but the most restricted budgets we’d recommend upgrading them to ensure your CPU achieves it’s potential, and to improve your quality of life through reduced noise. 

In this test, the ID-Cooling SE-224-XT shone through as the best entry-level cooler here. The combination of trouble-free fitting, good bundled accessories, and all-around performance means we are happy to recommend it to you for any build involving an entry to mid-level CPU. 

This test drove home the point that cooling capacity and noise are closely correlated to the size of the fan – we’d recommend opting for at least a 120mm fan cooler provided it fits in your case and you can afford one.

Whilst the Bequiet! Pure Rock Slim and Arctic Freezer 7X both acquit themselves well, without intrusive fan noise, they include compromises: the Bequiet! runs slightly hotter, and can’t be mounted facing front to back on AMD motherboards. The Arctic Freezer 7X has a frustrating fitting mechanism when used with Intel. At the right price, they’re a useful step up from the stock included coolers, once you’re sending $20-$25 we’d suggest finding a few dollars more for the ID-Cooling SE 224. The basic version is available for just $30.

If these coolers aren’t available, we’d recommend the following coolers: The Deepcool Gammax 400 is an excellent value choice, with four heat pipes and a 120mm fan it will perform much like the Hyper 212 EVO V2 and SE-224-XT. The BeQuiet! Pure Rock 2 Black is available at $44 and marries a single fan 4 heat pipe arrangement with a slick black aesthetic.

And whilst the Arctic Freezer 7X disappointed, the Arctic Freezer 34 E-sports uses a more versatile and conventional mounting system, and a 120mm fan. It’s available in a range of colors to match your build and is a good choice at around $35. Finally, The Noctua NH-U12S Redux creeps in at under $50 and uses 4 heat pipes and Noctuas very robust mounting system for a low noise alternative.

The post Best Entry Level CPU Coolers: Our Roundup for 2021 appeared first on PremiumBuilds.

When you’re looking to cool a more demanding CPU like the Ryzen 7 5800X or 5900X, or the Intel i7 or K series variants, you might want to consider a high-end air cooler. These CPU coolers offer the advantage of lower noise and increased reliability over ‘all in one’ liquid cooling systems thanks to their large fans and simple but effective design.

We decided to obtain a range of examples from the market leaders Noctua and BeQuiet! In order to assess them for compatibility, ease of fitment and take a look at the performance they offer – to give you a comparison of the best air coolers for high-end PC builds in 2021.

Starting out with more compact options we take a look at the Noctua NH-U12S, and the bequiet! Dark Rock 4. These coolers offer the best compatibility and clearance thanks to relatively compact designs. The NH-U12S uses a 120mm Fan whilst the Bequiet Dark Rock 4 uses a 135mm fan.

Then there’s the option of splitting the difference between the compact and dual fan coolers, the Noctua NH-D15S, which uses just one 140mm fan and a split heat sink – offset to assist in compatibility.

Finally, we pit the two dual-fan options against each other: the Noctua NH-D15 is the most powerful air cooler on the market, and the huge heat sinks and pair of 140mm fans promise great things. Up against it is be quiet’s Top offering, the Dark Rock Pro 4. Again with two fans but differing in size, this cooler blends great looks with low noise.

Best Air CPU Coolers – Our Recommendations

1. Noctua NH-U12A

The Noctua NH-U12A is a high-performance tower cooler with 7 heat pipes, a single fin stack, and a pair of 120mm fans in a push-pull configuration. It comes supplied with a wealth of high-quality fitting equipment, including thermal paste, fan splitters and ‘quiet fan adaptors’, and nicely produced instructions. There’s a robust metal backplate supplied for Intel.

Fitting is easy, with the exception of some confusion over the correct bushings to use: there’s white, beige and black bushing supplied, and it’s the black ones that are used for intel LGA1200 and LGA1155 fitment, but we had to check this with Noctuas online fitting videos as the instructions aren’t clear, mentioning a part number when stating the colour would have been far more helpful. The brackets are robust, and you can fit them horizontally or vertically according to preference. There are no compatibility issues with RAM, or the GPU on an mATX motherboard so it’s versatile, whilst the total height of 158mm means it’ll fit into a wide variety of cases.

Cooling performance is excellent as you’d expect and the fans are barely audible although they run at 1200 RPM, faster than the 140mm fan coolers in this test. This isn’t intrusive despite 120mm fans thanks to the exceptional fan quality. Temperatures are on a par with the larger coolers, sitting at 60C. Noctua suggests using one of the fans to replace the case exhaust, and in most configurations, we’d second that as a great low noise alternative to a supplied case fan. Moving from push-pull to a single fan barely affects temperatures at all. 

Any criticism of the U12S is either personal preference or splitting hairs: The instructions could clear up the bushing confusion with a single mention of the correct colour. The colour of the fans is entirely down to preference, but they are somewhat love or hate, and hard to integrate into a more showy build. There are chromax adaptors to cover the heat fin stack, but they’re extra. It also feels a little expensive, largely down to the quality of fans included which are unrivalled.

Ultimately, this cooler is all about performance, and it nails that. If you need a top tier solution with reliability and engineering a priority, and don’t mind the looks, then this is the cooler for you.

2. Be Quiet! Dark Rock 4

The Bequiet! Dark Rock 4 is a six heat-pipe heatsink with a single 135mm fan. This cooler is supplied with a bar-style retainer and high-quality bracketry, as well as a really high-quality long reach screwdriver that you can use for the rest of the PC build. Assembling the metal backplate and pins for intel applications is fiddly but other than that fitting is straight forwards. The fan does interfere with the first RAM slot for all but the lowest profile RAM, so if you’re planning four sticks you might want to look elsewhere, or else ensure the case has clearance for the fan to be mounted higher. This configuration negatively impacts the clean looks of this cooler.

The Dark Rock 4 is another largely silent cooler with good thermal performance. In our testing fan speed is amongst the lowest at just 800RPM, whilst it allows temperatures to sit a little higher at 70C. Adjusting fan speed allows you to strike the balance that’s right for you. It’s well suited to higher performance AMD CPUs, and it’ll handle Intel’s latest i5 CPUs, and the i7 non-K CPUs too. If you’ll looking at cooling an i9, the 10850K or 11900K for example, we’d recommend something with more cooling power as those CPUs can overwhelm even this cooler. For everything else, it’s recommended as a compatible, well made and good looking cooler with great reliability and performance.

3. Noctua NH-D15S Chromax.black

Straddling the gap from single fan to dual fan coolers, the Noctua NH-D15S is a split heatsink cooler with six heat pipes. It’s supplied with a single 140mm fan. The cold plate is offset to provide more versatile compatibility with GPUs. It fits on an mATX motherboard, unlike the larger NH-D15. There’s also a notch in the radiator fins to allow for higher RAM, although we’d advise checking RAM height for clearance (it should be less than 62mm) before purchase. 

As you’d expect given Noctua’s attention to detail, the bracketry is blacked out on this Chromax version, to match the cooler. Fitting is easy enough though we’d recommend working out whether you’re better off fitting the cooler when the motherboard is inside the case, or outside – this really depends on the access into your case around the cooler once fitted and it’s even more of an issue on the larger dual fan version. 

Under test, the NH-D15S produces excellent thermals sitting at about 62C CPU temperature, and as you’d expect, virtually zero fan noise with the 140mm fan at about 850 rpm under load. The NH-D15S an excellent cooler, and usefully addresses some of the compatibility issues of its bigger brother making it a great all-rounder performance cooler for a high-end CPU. 

4. Be Quiet! Dark Rock Pro 4

The Dark Rock Pro 4 sits at the top of Be Quiet’s product lineup. It uses a split fin stack with a 135mm fan in the centre and a 120mm fan at the front face to drive air right through in a push-pull configuration. The smaller improves compatibility with RAM but you are going to want to carefully check RAM height before ordering this cooler, it does overhang slots and will obscure RGB RAM. It will not interfere with GPU’s on ATX or MATX motherboards. 

The fitting equipment for the Dark Rock Pro 4 is high quality, and the fitting procedure is well explained but tricky. The included screwdriver is absolutely necessary to reach the screw positions through the facing plate. Securing the fan retention clips can be tricky if the case doesn’t allow good access, whilst if you choose to fit the cooler prior to fitting the motherboard, you can run into problems screwing down the upper motherboard screw or fitting power and fan plugs. Plan your build process carefully around this cooler. 

Once assembled the cooler gives one of the cleanest and most impressive appearances of any air cooler, and thermal and noise performance is top-notch. This cooler will handle anything short of overclocking the most demanding consumer CPUs, for which you’ll want a water cooler with a large radiator, or a custom loop anyway. Thermal performance matches all but the Noctua NHD15 in our test, whilst fan speeds remain low at 850RPM. It takes our recommendation due to the combination of performance, compatibility and low noise.

5. Noctua NH-D15

The largest air cooler in this test is the Noctua NHD-15. This is a split heat stack cooler but despite only one letter difference in the name from the NH-D15S, there are some considerable physical differences. The cold plate is not offset, meaning this cooler is only compatible with ATX motherboards – it obscures the top slot in an mATX motherboard making it impossible to fit a GPU. It’s supplied with 2 fans of equal 140mm size, so the front fan will interfere with RAM unless mounted higher, or at the rear of the cooler not the front. Fitting is straight forwards but careful planning needs to go into order of operations: You’ll need to fit RAM before the cooler itself, and ensure that you plug in motherboard plugs before fitting the cooler. Access to do up the fan clips can be hard too. You’ll need a large case and some serious planning to fit this cooler into your build. 

Once fitted, this cooler excels in performance testing: It’s silent and powerful. It kept temperatures the lowest of any cooler on the test, at 58C, with fan speeds matching the other high-performance coolers at 850rpm.

The downsides are really the cost, and the compatibility ‘cascade’ you need to consider to incorporate this cooler into your PC. The motherboard type, case size and ram selection all hinge on this Cooler. If you were planning a full-scale ATX build anyway that’s not a compromise, but it does preclude more tightly packaged builds. The looks are down to personal taste, but if you prefer a less distinctive colour scheme there’s also a ‘Chromax’ black version for $10 more. Noctua also supplies Chromax colour panels to adjust the scheme to match other aesthetics – though these do add considerably to the cost. 

This is undoubtedly the best performing tower cooler in this test, but you will need to plan your build carefully to incorporate it. It’s the go-to solution where reliability and performance trump other considerations, with redundancy provided by the dual fans. The other coolers, like the NHD15S or the Dark Rock Pro 4 come very close to the performance but are much more user friendly to integrate into a build and offer better value too.

Performance Testing

To test these coolers we mounted them to a test system comprising of an Intel i5-11500. This CPU consumes 110W at full load with power limits, giving sufficient heat output to examine the performance of these coolers. It’s equivalent to a Ryzen 9 5900X.

We can see that the Dark Rock 4 differs most from the other high-end tower coolers on test; It has the lowest fan speed at 750RPM and allows the CPU to reach the highest temperatures, but still delivers a very reasonable 70C overall temperature.

The Noctua NH-U12A is the only fan in the test with just 120mm fans, and the higher fan speed of 1200 rpm is the result of this. The cooling performance is still exceptional though and the very high-quality fans stay effectively silent.

The Noctua NH-D15, NH-D15S and the Bequiet Dark Rock Pro 4 are hard to distinguish – all deliver temperatures around 60°C and fan speeds at about 800RPM. All have ample headroom to tune temperatures or fan noise to your preference even with demanding CPUs. 

Conclusion: Take your pick on looks, value or compatibility.

Each of these high-end air coolers has much to recommend. The right choice for your build really relies on your preferences. The Noctua NH-D15S is impressive in performance and noise and is much more forgiving of RAM and Motherboard form factor than its larger sibling. Personally, we prefer the Dark Rock Pro 4’s combination of looks and thoughtful design, meaning it requires fewer compromises overall and looks fantastic in most systems without spending more on aesthetic accessories. 

Meanwhile, the Noctua NH-U12A and Dark Rock 4 absolutely have a place in build with slightly less demanding CPUs or more size constraints, you don’t sacrifice anything in noise or performance in opting for one of these CPU coolers. 

Lastly, the Noctua NH-D15 will find its home in a full-scale ATX build where high-performance CPUs need to work at optimal temperatures with minimal noise. If you’re using a large case anyway, this cooler brings AIO levels of performance without the complexity of adding an all in one water loop to your PC: It’s a great choice if you value low noise above all else.

The post 5 Best Air CPU Coolers for 2021 appeared first on PremiumBuilds.

In our review of the best B560 motherboards, we found a number of anomalies in their behaviour. In this article we’ll document what we found, to help you understand the problem and avoid the pitfalls. Some boards can be fixed by paying attention n to their settings in BIOS. Others cannot be saved and should be left on the shelf. So lets’ take a look at the potential pitfalls of the B560 motherboard range.

One of our first tests was a simple Cinebench R23 run, but given that the CPU and configuration were the same it yielded some surprising results:

The MSI Bazooka delivers 1500 points lower score than the two ASUS boards and the Gigabyte Aorus Pro. The ASock B560M HDV is about 800 points down on optimal CPU performance.

Let’s look at what happens if we dig back into BIOS and remove the power limits:

Clearly, something is up here with the ASROCK HDV and the MSI B560 Bazooka. The Bazooka now matches the other boards on test, the ASROCK performs better but still falls a couple of hundred points short of optimal performance.

To work out what’s going on, we compared logged metrics throughout these runs at default settings, specifically power consumption and CPU clock frequencies over time:

This graph clearly illustrates the differences in behaviour between these three boards. Here the solid lines represent core clock speeds, the dashed lines the power delivered to the CPU. The ASUS maintains power of about 110 watts throughout the test, keeping core clocks at 4200MHz. This is the result of ‘Asus multi-core enhancement’ being on by default, although the initial BIOS screen does advise you of this, I don’t expect its explanation to mean much to most users, except it seems like a good idea to leave it on.

The ASROCK board delivers 100 Watts throughout the first portion of the test and holds 4.2GHz all-core clock speed but then drops to 65W the long term power limit after about 50 seconds. This has the effect of cutting core clocks to 3.5GHz and results in longer test duration, and lower score.

Finally, we see the MSI Bazooka: Like the ASUS this initially delivers 110 Watts, and 4.2GHZ clock speed, but ramps down to 65W at just 24 seconds. Thereafter it holds a significantly lower 3.3GHz clock speed for the remainder of the test, finishing last and delivering the lowest score. This is a function of the ‘stock cooler’ settings being applied in BIOS.

This behaviour is the results of different implementations of Intel’s power specifications. Short term power should allow for 125W even on this i5-11500 which has a nominal TDP of 65W – that’s its long term power limit. We see it draw 110W on most of the boards because that’s the power required to achieve its maximum allowable all-core boost clock.

One of these graphs is more insidious than the others, and it’s not the MSI. If we adjust the power limit behaviour through MSI BIOS by selecting ‘Tower’ or ‘AIO’ cooler the MSI performs in line with the other boards. For the ASRock HDV, this means setting the power target, but it won’t allow user input of 125W, just 100W. Why is that?

This graph shows us why: We ran Cinebench for a 10-minute loop with power limits set as high as possible on the Asrock HDV, and the i5-11500. Here is power delivery vs Clock speeds over the duration of that test, 8 loops of the rendering task:

For the first three runs, all appear to be relatively normal, except that the CPU is only getting 100 Watts when it will use 110W for peak performance. By the fourth run we start to see something different though: The board spikes, then cuts power. Core clocks are no longer held at 4.2GHz but instead begin to fluctuate below that. As the runs repeat we see increasingly ragged behaviour. This VRM is failing to deliver clean power even at a reduced 100W level and is throttling the CPU as a result. Performance suffers. In the long term, if subjected to this workload, it’s clear the VRMs are over-extending themselves. And in case you’re thinking: ‘Surely that’s CPU temperature throttling’ The raw data shows that no core exceeded 60C throughout this test, and no core recorded thermal throttling at any point.

So, this is behaviour with a 65W rated part, an entry-level i5 CPU.

i9-11900K Testing

What happens is you dare to put a more power hungry chip on these boards?!

To test, we installed the i9-11900K on each of these boards, using unlimited power settings, to see what they were capable of. For comparison we’ve included the result this CPU gives on the ASUS ROG Maximus XIII Z590, a motherboard with insane VRMS capable of supplying well over 330W to this chip under demanding loads.

Here we can see the consequences of weaker VRMs. Both the MSI and the Asrock HDV, the two cheapest boards in this test, deliver substantially sub-par results. The Gigabyte and two Asus boards both achieve ‘full’ performance for this demanding CPU, nearly matching the Z590 at default behaviour.

Again, looking at the metrics, we can see how default board behaviour has a drastic impact on performance:

The Gigabyte is the only B560 motherboard that delivers sufficient power to reach a 4.7GHz all core speed, at 170W consumption. It maintains this for the duration of the test and records a score of 15,000 representing the potential of this CPU. The MSI initially ramps to 170W deliver as well but fails to lift core clocks beyond 4.2GHz and quickly falls back to a 130W power delivery, PL1. From there it maintains 4.2 GHz of the duration of the test.
Finally, the ASRock HDV starts out delivering about 135W for 18 seconds, then drops to the 65W PL1 limit. That means that the clocks fall back from 3.7GHz to 3.1GHz where they remain until the workload is complete. The ASRocks score in this test is just 10098. And as a reminder, that’s BELOW the score the six-core i5-11500 can post in this test given adequate power. Simply put: The CPU can only do as much work as you supply it power for. This ASRock motherboard limits that power enough to make the 8 core i9-11900K perform the same as a six-core i5-11500 in an all core workload, and if you don’t adjust power limits in BIOS, it’ll even make the i5-11500 drastically underperform. If you do lift those power limits and then subject the i5-11500 to demanding but entirely reasonable workloads, the VRMS can’t keep up.

Conclusions and the root cause

Fundamentally, the blame lies with Intel. This kind of underperformance due to power limitations is down to their inability to make the Rocketlake CPUs more efficient. Remember, none of the Ryzen Zen 3 CPUs, not even the 16 core 5950X demand more than 125W for full performance. Here we have a six-core part needing 115W to perform to full potential and motherboards failing to deliver the 170Watts that the 8 core i9-11900K requires to reach its all core boost clocks.

Then there’s the confusion around power limits. These CPUs quote boost clocks that are only achievable if you can deliver sufficient power and keep them sufficiently cool. Intel’s power specifications quote maximum power delivery and time periods – but there don’t appear to be any consequences for not adhering to them, breaking them, or simply not achieving them. Ultimately, when your boosting mechanisms are as complex and varied as Intels, you’re not setting a clear picture of the level of performance consumers can expect from your products. 

A mid-range CPU that can exceed 100 Watts draw in an all core load poses a significant problem for motherboard manufacturers. The i5-11400 has been touted as the new value champion for gaming – but you need a motherboard to run it. It’s sold with a 65W TDP but that’s not an accurate representation of its power demands. So motherboard manufacturers want to cater to a value-conscious market but need to make a board that can potentially be fitted with any CPU from an i3 up to an i9-11900K, or more realistically perhaps an i7-11700.

ASROCK has abjectly failed to do that. The B560M HDV is plainly inadequate. It throttles an i5-11500 at base settings. It will not allow those settings to be lifted to a point that achieves full performance from the CPU. It cannot sustain an all core workload for ten minutes without the cracks showing through. It’s not acceptable. 

Then we come on to the wider issues: Across the boards, we’ve sampled we’ve seen wildly different behaviour because of the power limits set and implemented by manufacturers. Credit goes to Gigabyte and Asus for manufacturing boards that can achieve the full potential of even demanding CPUs – but they’re not actually adhering to Intel’s specifications either, by allowing power limits that exceed specifications and durations that also run longer than those specifications. MSI are at least open about the need to define power limits when you set the motherboard up.

Some of these boards allow you to set your own power limits to achieve your desired result, but we don’t believe that consumers should have to do that on motherboards that are aimed squarely at the mainstream. If you’re buying a B560 board you have every right to expect it to work well with an i5 or i7 non K CPU without manually adjusting settings or understanding the detail of power delivery and limits, but some of them don’t.

So my ultimate conclusion is that if you do want to make use of an i5-11400 or any other Intel CPU and you’re considering a B560 board, it will take some research and sadly a little more money to get the most of your CPU. You can see out the linked article for the best B560 Motherboard recommendations, in tandem with the information presented in this article.

The post B560 Motherboard Power Limit Analysis: A Real Minefield of Bad Boards, Confusing Settings, Tricky Choices! appeared first on PremiumBuilds.

The Intel Rocket Lake i5 CPUs represent some of the best value gaming options on the market right now, particularly the i5-11400. To get the most out of them you need a B560 motherboard, so we thought we should test them out so we can make some recommendations to you.

B560 Motherboards under test

There’s a lot more going on here than simple features and looks – there’s some great B560 motherboards here but there’s also a product you’re definitely going to want to avoid.

We’ll take you through the features of these motherboards, how easy they are to install and set up, the BIOS configuration of each, and give some insights into performance because unusually some of these motherboards can actually limit the performance of a CPU, even an i5-11400 – one of them because it’s just a bad B560 motherboard, others if you don’t set them up correctly. Because we’ve selected boards across the range and from different manufacturers, we can use this to recommend other boards not on this test but that share components or specification with them. We’ve used our testing to give recommendations for boards to pick for the entry-level, mid-range and at the high end, so let’s get stuck in!

1. Features Overview

All of these boards share some basic features common to the B560 platform. All are PCIe 4.0 compatible in the primary M.2 slot and PCIe slot. All have 2 M.2 slots total. They all allow RAM overclocking.

The ASRock B560M-HDV is unquestionably the bottom of the stack. It has just 2 RAM slots, and it’s cut down from the normal mATX size using just 6 mounting points instead of the usual 8. It doesn’t have any POST code troubleshooting lights or a Bios flash button. The rear IO lacks Displayport and instead has HDMI, DVI and even a D-Sub port, along with 3 audio jacks and just 6 USB ports. There’s Gigabit LAN but no WiFi. It has just one full-length PCI-E x16 slot and two x1 slots. It lacks a USB 3.2 gen 1 header so many more modern cases aren’t compatible without an adaptor, and will lose that connectors potential speed. It has just 4 SATA ports for additional drives. There’s no heatsinking on the m.2 drive or the Voltage delivery circuitry – but more on that later.

The MSI B560M Bazooka has heatsinking on the VRMS and primary M.2 slot, and a military-themed aesthetic. There are 4 RAM slots but still just one full-length PCIe slot and two additional single length slots for WiFi cards or similar expansion. The rear IO is pretty sparse, with just 6 USB ports, 3 Audio jacks, and a 2.5Gigabit Ethernet port. There are HDMI and Displayport outputs if you want to use the iGPU. It does have some basic troubleshooting boot LEDs and 6 SATA ports. There’s an AIO pump header and adequate fan headers for most builds. It does lack the USB 3.2 Gen2 header so check your case compatibility. There are RGB headers but no RGB on the board.

The Asus TUF Gaming B560M-Plus WiFi plus is firmly mid-range and has a good suite of features. Importantly it has inbuilt Wifi 6, so there’s no need for an additional card to get WiFi and Bluetooth connectivity, it’s right there on the rear IO. It’s got 8 USB ports on the rear plus a USB C, five audio jack connectors plus optical out, and 2.5 Gigabit Lan, as well as Displayport and HDMI.  The IO shield is inbuilt too, making install a cinch. There are 2 full-length PCIe slots so a capture card or similar ‘X4’ secondary card will fit in the lower slot. There are 4 RAM slots, 6 SATA ports, and a USB 3.2 Gen 1 header to give you those high-speed ports on the front of the case. It has the ‘TUF’ militarised aesthetic and some neat RGB highlights on the board. 

The Gigabyte Aorus Pro AX represents the higher end, with a sleek silver aesthetic, heavy aluminium heatsinking and an integrated IO plate. They’ve crammed in 9 USB ports on the rear, and a USB C socket as well, alongside the WiFi 6 antenna connectors. Internally there’s 2 full length PCIe slots, and one awkwardly placed single length slot, along with 6 SATA ports and a USB 3.2 gen 1 header socket. There’s also a good number of USB 3.0 and audio connectors as well as RGB control headers. It has a Q-Flash button on the bottom edge of the board, useful for recovering BIOS or updating without RAM and CPU installed. It’s worth mentioning the VRMs here, this board has a 12 phase 50 Amp VRM set up which is clearly the best specified in test.

Finally, we come to the only mini ITX board on test, the ASUS B560-I ROG Strix. This is a premium board but contends with the constraints imposed by its tiny size. The rear IO is integrated and includes 6 USB sockets, a USB Type C, WiFi 6 antenna connectors, and five audio jacks and unusually it also has an audio type C connector. There’s 2.5Gigabit ethernet, a Displayport and HDMI output. Internally there are two M.2 slots, one front at PCIe 4.0 speeds with dual-sided heatsinking and a rear PCIe 3.0 slot as well. You’ll have to contend with the mini ITX limitations of a single PCIe slot, 2 RAM slots, 4 SATA ports, one each of USB 3.0 and 3.2 gen 1 header, and one each of CPU, Chassis and AIO pump fan headers. ASUS include a handy USB C extension cable – and a key ring…

2. Ease of set up and install

Full marks go to Asus here for their M.2 clip which makes drive installation a cinch. The manuals are clear and accurate, and they have integrated backplates to ease installation.

The Gigabyte also has an integrated IO plate, but it loses marks for an awkward M.2 installation solution with a split standoff. The heatsink looks like it’s hinged but it’s actually just a tab. The awkward process is exacerbated by a manual that doesn’t have clear diagrams or even reference the m.2 installation process. It’s outdated and could do with a refresh to help first time builders along.

The MSI Bazooka has a separate I/O backplate, but the m.2 installation is easy enough and the manuals are clear. POST lights help diagnose any installation or settings errors. 

In a clear last place is the ASROCK HDV. The IO shield is incredibly basic pressed tin, there’s no diagnostic lights or BIOS flash button to help troubleshoot. At least with no M.2 heatsink, and only 2 RAM slots, it’s hard to get it wrong when you put it together. There’s a major problem when it comes to the configuration of this board though, which we’ll come to in the performance section. 

3. BIOS

You don’t often see a critique of BIOS in reviews. It’s a one time deal for most people who just want to build and set up their PC and won’t re-enter BIOS unless something goes wrong.

B560 is the first non-enthusiast chipset to permit RAM overclocking, so that opens up the need to enter BIOS to set up RAM correctly, and delve through menus and settings that otherwise you may never see. There’s another issue here, and that’s that some of these boards have default power settings that actually limit performance even of i5 CPUs. That means you might need to dig into power settings to get full performance and if the BIOS isn’t intuitive that can be a tricky task.

BIOS is somewhat subjective because your comfort with a particular lay out hinges on what you’re used to. However, in our opinion, the MSI BIOS is the best laid out and most intuitive to use. On initial set up it’s the only BIOS that clearly explains the need to set your power limits according to the cooler you’re using and shows exactly what the limits will be set to. Once you’re in BIOS, there’s a basic simple mode that lets you select XMP and boot order, likely the only 2 tasks you’ll need to do. Advanced mode opens up a wealth of options including fan tweaking, Memory overclocking, setting power limits for the CPU, and saving and applying profiles. 

Asus and Gigabyte suffer from the same slightly confusing advanced layout, with numerous options jumbled together and not always in a logical order. Asus buries sub-menus making them hard to navigate. It does however default to using ‘Multi-core enhancement’ which allows a CPU to utilise higher power limits, but it’s not entirely clear to the first time user exactly what this means and it’s very much sold as an option you should select. 

Gigabyte makes no mention of power Iimits: It appears the option may depend on what CPU is fitted. It suffers from the same problems as ASUS with a slightly confusing BIOS layout but does make some attempt to explain some settings like Load Line Calibration which is nice to see.

The ASROCK BIOS is acceptable in layout and function, but again the power settings are both obscured, and actually do not allow a free choice of power limits, likely because of constraints of the motherboard’s design. As a result, you cannot set a power limit higher than 100W and this has a negative effect on performance. Other than that, the Asus, Gigabyte and ASRock BIOSes are similar in layout and function and serve their purpose well enough. 

4. Performance

Firstly, one of the main reasons the B560 chipset is recommended is because it allows memory overclocking for the first time officially on non-K CPUs. This can bring decent performance benefits and is worth exploring if you are opting for this platform.

All of these boards allow memory overclocking, all have very similar settings available to enable that, and all allow running memory in ‘GEAR 1’ mode at 3600MHz – meaning the memory controller clock matches that of the RAM itself. All the boards accepted the XMP profile on the 3600MHz CL16 G.Skill kit we used for testing. 

Where we start to see some important differences is in how the boards behave due to their implementation of Intel’s power limit specifications, especially at default. 

To demonstrate, here’s a graph showing the Cinebench R23 score of each of these boards running an i5-11500 CPU using default power settings:

Whilst the two Asus boards and the Gigabyte Aorus all turn in similar results at about 10,200, we can see the MSI Bazooka and ASRock fall significantly behind. What’s going on?

Logging metrics, we can compare the CPU behaviour through this test on the different B560 motherboards.

This graph shows the impact of power limiting behaviour by the motherboard. The MSI deploys it’s high power limit which allows the CPU to draw 110W for the first section of the test, and permits running the CPU at it’s rated 4.2GHz all core speed. It then drops to a 65W long term power limit which limits clock frequencies to 3.3 GHz and causes a low score.

The ASRock HDV initially appears to be doing better: But look closely: It is never able to deliver more than 100W, and it then drops to 65W and the clocks to 3.5GHZ.

Meanwhile, the Asus TUF maintains 110W for the full duration of this test, running the CPU at 4.2GHZ throughout, and that’s down to ‘multi-core enhancement’ being active by default.

The MSI Bazooka’s behaviour can be fixed: Selecting ‘Tower’ or ‘All in one water cooler’ on initial set up in the BIOS imposes a higher power limit and allows the CPU to perform to its full potential. 

The ASROCK HDV however, cannot be saved. Imposing a 100W limit in BIOS improves the CPU performance as can be seen in this re-test with power limits removed, but its performance still falls short of the other B560 motherboards. If you’re thinking ‘hey, that’s not too bad, it’s only 200 points behind’… all I can say is please watch our linked video exploring this issue on B560 boards. We’ve got much more testing including testing with an i9-11900K, and they show why this is such a bad result for the ASROCK HDV, and why you shouldn’t consider this motherboard for even an i5 build. 

In terms of performance, the ASUS, Gigabyte and MSI boards perform equivalently and are able to extract the full potential from this CPU whilst the ASRock HDV gets a serious black mark against it for failing to meet Intel specification and failing to extract the full potential from an i5 CPU. 

Conclusions: Power delivery matters on B560 Motherboards

Our testing has highlighted important differences in the B560 chipset motherboards available. You need to factor in not only features and price, but also potential performance issues with some of these B560 motherboards.

The performance issue really is one of Intel’s own making: Their dogged adherence to squeezing every last drop out of aged architecture and process leaves them with mid-range CPUs that draw a lot of power when performing at their full potential. These components aren’t cheap, and this leaves motherboard manufacturers in the unpleasant position of trying to deliver compatible boards at an attractive price point. It appears that boards like the ASROCK HDV or some of the Gigabyte UD offerings are intended for use with i3 CPUs in office style low-demand builds because they’re not suitable for more demanding CPUs or usage. 

B560 Motherboard Recommendations

Best Entry Level B560 Motherboards

In our testing, the MSI MAG B560M Bazooka proves itself to be an excellent entry-level B560 motherboard. Or, if your budget is tighter, the MSI Pro VDH which is identical in specification and has VRM heatsinks, but loses the gamer aesthetic, but is also available with inbuilt WiFI.

ASUS has used a robust 8 phase VRM design across their affordable Prime Range, and also have heatsinking. The B560M-A or B560-PLUS are also good options at the entry-level price point.

Best Mid Range B560 Motherboards

The MSI MAG B560M Mortar and Asus TUF Gaming B560M-Plus WiFi both hit the sweet spot of solid VRM design and features and both have integrated WiFI 6 options available.

If you are looking for an ATX board the MSI MAG B560 Torpedo is a strong option with similar specification to the Mortar.

Best High End B560 Motherboard

Gigabyte has done a good job of VRM design since the ‘Coffee Lake’ era, and the VRM on the B560M Aorus Pro AX is no exception. The closely matched Aorus Elite and Aorus Pro are near identical in specification, with the Pro version having slightly stronger components in the VRM but retaining the same basic design. Either B560 motherboard will make a great choice for an i5 CPU. The Aorus Pro is our pick for the strongest VRM on test, it had no problems supplying power to the i9-11900K and its all-round specification matches the other motherboards on test.

Beyond $200 – Look to Z590 or AMD Zen 3

If you’re pushing towards $200 you perhaps look to a Z590 motherboard instead. Whilst the Asus Strix and MSI B560 Tomahawk are good offerings, you should look to gain the flexibility of a Z590 motherboard as you approach that price point. 

You could also consider the platform cost of a switch to a Ryzen Zen 3 CPU and a more cost-effective AMD B550 motherboard, where there’s a range of good options for around $100-$150 that offset the slightly higher CPU cost. There isn’t really a lot of sense in trying to get the value out of a CPU like the i5-11400 or i7-11700, but then overspending on motherboards to support them.

Best Mini ITX B560 Motherboard

If you’re looking for a mini ITX board, then the ASUS ROG Strix B560-I doesn’t disappoint. It allows both the i5-11500 and i9-11900K to perform to potential and has the same features as any other B560 ITX board on the market. You might want to consider the heat loading of an Intel Rocket Lake CPU in a small form factor build, especially when compared to a more efficient AMD Zen 3 CPU. That said, it’s a great looking and well-specified board and gets our recommendation from the handful of mITX B560 offerings.

That concludes our roundup. We hope you have this interesting, and perhaps it’s saved you from buying a part that would lead to the disappointing performance of your build. The B560 motherboards we’ve recommended will provide an excellent basis for your next PC, with a great blend of features, performance, ease of use and good value too. 

We also have produced a video companion to our B560 motherboard article, which we have linked below.

The post Best B560 Motherboards for Intel Rocket Lake Builds appeared first on PremiumBuilds.

In this article, we’ll investigate the impact of RAM speed on the Intel Core i9-11900K and i9-10850K in gaming. It’s well known that Ryzen Zen 2 and Zen 3 CPUs perform optimally with higher speed RAM, but we’ve often heard it said that ‘RAM speed doesn’t matter as much with Intel’. We wanted to test that. 

With the 11th Gen Rocket Lake CPUs Intel introduced the notion of ‘Gear 1 and Gear 2’ settings for the memory controller. This is analogous to Ryzens’s uClk setting, in which it’s important to match the memory controller clock and infinity fabric speeds to RAM speed for optimal performance. Now, Intel has also handed control of this behaviour to the end-user – or the motherboard manufacturers – so you can adjust this behaviour in 11^th generation K series CPUs. 

We’ll start by looking at RAM speeds and controlling Cas Latency, then move on to looking at Gear 1 vs Gear 2 and what that means for performance. 

The Test Setup:

Whilst testing speeds we used the ASUS ROG MAXIMUS XIII Hero Z590 Motherboard with the recent 0605 BIOS which included Rocket Lake-S specific microcode and improved memory stability and flexibility. We kept CL timings fixed at 16-16-16-32 except for the higher clock speeds where loser timings were required for stability. We used our 16Gb Samsung B-die RAM kit, capable of 4400MHz speeds but changed RAM settings to the primary timings and speeds shown.

i9-11900K & i9-10850K RAM Performance Analysis

Shadow of the Tomb Raider: i9-10850K

Looking first at Shadow of the Tomb Raider on the Intel Core i9-10850K you can see how marked the trend is.

This is the benchmark reported CPU performance, independent of GPU. Moving from 2400MHz CL16 RAM to 3600MHz with all other settings controlled yields a 30 FPS performance increase. Further increasing speed to 4000MHz and reducing latency, but by a smaller amount owing to the lift in CL timings, gives another 4% or 8 FPS average. At 4400MHz CL19 we see our highest performance at 195 FPS average, which’s 43FPS or 28% faster than the base settings. To be clear, 2400MHz or even worse 2133MHz is what will happen by default if you fail to set XMP or the XMP profile on your RAM kit isn’t stable and it resets itself. It can cost you significant performance.

Here’s Rainbow 6 Sieges benchmark, a fast-paced shooter where every FPS matters. Again, this benchmark is consistent and RAM speed-sensitive.

Here we get a slightly different trend, with the best performance at 3600MHZ CL16, and then performance softens at 4000MHz CL17 and 4400MHz CL19. Again, here our overall latency and the looser timings required to keep RAM stable at those very high speeds becomes detrimental to performance.

Shadow of the Tomb Raider: i9-11900K

Moving on to Intel’s latest CPU, the i9-11900K and using the same benchmarks, we can see the same trend with a couple of caveats. Note these tests were performed in Gear 2 mode because Gear 1 was unstable at the higher memory speeds. That’s also why 4400MHz is missing from these results, it wasn’t possible to stabilise it in the gaming benchmarks at that speed. 

Here again, we see the serious boost in performance moving from 2400MHz up to 3000MHz, then to 3600MHz adds a total of 55FPS average to the CPU performance, with corresponding hikes in minimum and maximum performance metrics. There’s a 41% boost to performance moving from 2400MHz to 3600MHz CL16.  At 4000Mhz Cl17 performance tails off marginally – again we’re approaching the limits of the memory controller and slacker secondary and tertiary timing dent performance overall. 

Confirming this behaviour in Rainbow 6 Siege, working through the same settings yields the following results:

Again we see peak performance centred around 3600MHz CL16, and a softening towards 4000MHz but were not giving much away between 3000Mhz and 3600Mhz with reasonable timings, 10 FPS or so.

And finally, just to confirm this trend and to show a little data for 4400MHz RAM on the 11900K, we were able to get the Time Spy CPU benchmark to complete at 4400MHz CL19, building the following set of results:

This is perhaps less informative aside from the fact of having that highest speed represented as it’s just a score, but it is indicative of aspects of the CPU performance and we know from another testing that this benchmark does scale well with RAM latency. It looks like we’re really at the bitter end of the memory controllers performance at these higher speeds. It would take a significant investment of time in refining RAM timings manually to, first of all, stabilise the system, and further improve performance. Nevertheless, 4400MHz CL19 does yield the highest score here but it’s not stable enough to run any games.

Gear 1 vs Gear 2

Moving on then to the question of Gear 1 Vs Gear 2 on capable Rocket Lake 11th Generation CPUs, this is a simple toggle in BIOS that runs the memory clock at either half memory speed in Gear 2, or at the same speed in gear 1. Gear 1, therefore, reduces latency, as there’s no missed cycle in communication. The penalty for this is the memory controller becomes less flexible in settings, so we’ve got limited data. Like Ryzens uCLk and FCLk settings, it also appears to be ‘unhappy’ at higher frequencies, and it wasn’t possible to stabilise behaviour at 4000MHz RAM speeds/2000MHz memory controller speeds in the time available. We satisfied ourselves with conducting A-B testing to see what kind of performance difference we could see at settings we could apply consistently across the tests. 

Here we can see small but significant performance benefits from running at Gear 1 and 3600Mhz, with a 10FPS uplift in average frame rates on Rainbow 6 Siege, and we verified this uplift at 2400MHz base settings as well although of course, overall performance is much lower due to the overall speed impact discovered above.

Finally, for a real-world demonstration of this benefit, Flight Sim 2020 is a highly demanding simulator that is frequently CPU bound. Running our standardised benchmark flight over Manhattan yields the following results:

Here, we can see that moving to Gear 1 and matching memory and memory controller speeds yields a small but valuable lift across the full suite of metrics and translate directly into improved performance.

Verdict

This set of results demonstrates how important overall memory latency, as indicated by speed and timings, is for Intel CPUs. Clearly, it can have a significant performance impact in any situation where you’re wholly or partially CPU limited. 

It highlights the importance of ensuring that your Memory is correctly set up. If you didn’t enable the XMP profile on your RAM during set up, or it’s been reset in a BIOS update or by clearing your CMOS, you could be hurting performance significantly. It takes a moment to check this using CPU-Z or Hwinfo64 and doesn’t necessitate entering BIOS. 

For both Comet Lake and Rocket Lake K series CPUs, there appears to be a sweet spot in the region of 3600MHz CL16, and that’s a good starting point for refining system performance. To clarify the settings we used here were ‘quick and dirty’ because of time constraints, with the motherboard left to decide most secondary and all tertiary settings itself. Time spent tightening those timings could yield further performance improvements, but it’s a time-consuming process and tends to offer diminishing returns, and you have to check and validate results not only for performance but for stability as well. 

For most users with K series CPUs, it seems that much like Ryzen you’re best off buying a kit of 3600MHZ CL16 RAM or something with a similar total latency, and if you do that and enable XMP you’ll have gained the bulk of the performance benefit for virtually zero effort. If you spend as much time in BIOS as your operating system, then buying even faster RAM and using the headroom to lower latency as much as possible can yield significant rewards on these CPUs.

The post Rocket Lake RAM Speed Analysis: Best RAM Speeds for Core i9-11900K (and i9-10850K) appeared first on PremiumBuilds.

Intel’s brand new flagship, the Core i9-11900K releases today and we’ve been given one to test and review. In this article, we’ll put it through its paces against its closest competitor in specification: The AMD Ryzen 7 5800X, and we’ve got an Intel Core i9-10850K for comparison as well because it’s the current high-performance value champion.

Intel has been lagging behind in the CPU wars for six months now. They’ve lacked a CPU that can challenge AMD’s Zen 3 line up for raw performance and were missing features notably PCIe 4.0 support. The new 11^th generation Rocket Lake CPUs seek to address that Intel is making some bold performance claims.

This i9-11900K CPU boasts 5.3GHz peak boost speeds using Thermal Velocity Boost, 8 cores, and 16 threads. It uses the new Xe architecture integrated GPU. The road to Rocket lake hasn’t been smooth though: it’s suffered a convoluted development, originally scheduled to be released on a 10nm production process, then backported to 14nm when that failed. This the end of the line for this architecture, this process node, and this socket as far as Intel are concerned. This should represent the pinnacle of their current capability so we’re eager to find out what it can do.

Test methodology

We’ve taken great care to ensure this test is fair. To do that we’ve controlled every variable that we can. All the synthetic and gaming results you’ll see are obtained with the same RAM settings across the 3 CPUs under test. We’ve tested using an up to date BIOS (0605), released just 6 days before this release. We’ve used exactly the same motherboard for both Intel CPUs, and a B550 motherboard for the Ryzen 5800X, the MSI Mortar. 

For all the gaming and synthetic tests, we kept to Intel’s specifications for multi-core enhancements, power limits, and Thermal Velocity Boost. We did this because to our mind this is comparable to how we’ve tested the 5800X using PBO. Both CPUs were allowed to perform as they do with minimal setup, according to the manufacturer’s intentions, but with the automatic optimisations in place. It’s also the default behaviour of this motherboard. 

We verified this behaviour with A-B testing in a number of metrics and with both our RAM settings and motherboard settings the results represent this CPU performing at its best, outside of more involved manual tuning or overclocking. RAM was set to 3600MHz CL16-16-16-32 in all tests except the specific memory tests. There’s also the issue of ‘Gear 1 and Gear 2’ memory controller settings analogous to Ryzen’s Infinity Fabric and memory controller ratio settings – these tests are run in Gear 1 with the memory controller speed matched to memory speed. We’ve also got a separate article digging deeper into the impacts of Memory speed on performance on this CPU.

The Test System

We ran both Intel CPUs in the Asus Z590 ROG Maximus XIII Hero. With 14 phase 90Amp VRMs this high-end Z590 motherboard is an overclockers dream and we found it very flexible in terms of memory settings. We ran the tests with the 0605 BIOS from ASUS including Intel’s latest Microcode updates. We used a Fractal Design Celcius S28+ AIO Cooler and an Ion+ 860W Platinum Power supply. 

For RAM, we used our 16GB Samsung B-Die 4400MHz CL16 kit, but run it at 3600MHz CL16 in order to match as closely as possible the settings in our Ryzen testing.

For the GPU we used the EVGA RTX 3080 XC3 Ultra but run our test settings in order to expose the CPU performance as much as possible, this powerful and consistent GPU helped us do that. 

The Ryzen comparison system is identical with the exception of an MSI Mortar B550 Motherboard.

Specifications

CPU	Intel Core i9-11900K	Intel Core i9-10850K	AMD Ryzen 5800X
Design
Price	$539	~$380	~$449
Cores/Threads	8/16	10/20	8/16
Process	14nm	14nm	7nm
Architecture	Rocket Lake-S	Comet Lake-S	Zen 3
Peak boost	5.2GHz	5.2GHz	4.7GHz
Boost Technologies	TVB, Turbo Boost Max 3.0, Adaptive Boost Technology	TVB, Turbo Boost Max 3.0	PBO
On Board Graphics	UHD 750	UHD 630	None
Power Draw	125W (TDP) ~250W Max	125W (TDP) ~250W Max	105W (TDP)

Intel Core i9-11900K Performance Analysis

Synthetic Benchmarks

Cinebench R20

Cinebench R20 is a test of single and multicore performance whilst rendering a scene. It is almost entirely independent of memory speed which allows us to isolate raw CPU performance. 

We conducted three runs and averaged to obtain these results. In the multi-core tests, we see that the 10850K and 5800X are neck and neck, both ahead 100 points at 5990 to the 11900K’s 5860. The 4 point difference between the 10850K and 5800X is imperceptible but the last-gen Intel CPU has 10 cores, not 8 to obtain this result.

Looking at the single-core performance again averaged over 3 runs we can see the difference: Both the 11900K and 5800X score an identical 624 points average, whilst the 10850K lags 100 points behind with a score of 516. 

This result is close enough that cooling set up or silicon quality on the chip could influence it but on the raw numbers, the Ryzen 7 CPU performs best in Cinebench R20 overall by matching the 10850K’s multi-core score, and the 11900K’s single-core score.  

Blender

Using Blender to render a couple of scenes, again we get a sense of the rendering performance of these CPUs.

Note that shorter bars are better indicating less time taken: Different scenes favour different aspects of a CPUs performance, and in this test, we can see that for the ‘Classroom’ render, the 11900K and 10850K are neck and neck at 435 seconds, but the 5800X finishes first about 20 seconds quicker.

In BMW27 the i9-10850K takes the lead at 135 seconds, the 11900K is 10 seconds slower, and the 5800X finishes last about 20 seconds behind. There’s no clear winner here, and I feel obliged to point out that we’re using this as a test of the CPUs, and if you’re actually looking to accelerate 3D rendering a GPU will complete the task in a fraction of the time of these CPUs.

3DMark

Moving on to gaming-oriented benchmarks, 3D Mark. Focussing in on the CPU component of the Fire Strike and Time Spy benchmarks, these tests do bring memory performance into play somewhat and also heavily favour higher core counts as it’s a parallel test that uses all cores. 

The i9-11900K places last in Fire Strike, 500 points behind, and splits the difference between the other two CPUs under test in Time Spy. 

So rounding out our synthetic benchmarks, we see a picture of the i9-11900K having a high single-core speed, on a par with the 5800X and able to match the 10850K in some workloads despite having 2 fewer cores. But it’s not faster and struggles to make a mark in these tests.

Game Benchmarks

We ran our gaming benchmarks at 1080p to isolate CPU performance as much as possible, but retained settings that are relevant in the real world. The RTX 3080 helps us see differences in underlying performance. 

Call of Duty: Warzone

Call of Duty: Warzone is our first test, and we ran a 5-minute battle Royale against bots to try and give an overview of performance, not a snapshot. This game is a mix of CPU and GPU performance and you need both to really achieve high frame rates even at 1080p.

We can see that the 10850K and 11900K perform almost identically here, within a couple of FPS on Average scoring just over 200FPS, min and maximum metrics. The 5800X is the clear winner though with stellar performance and 240FPS average. It’s disappointing that we’re not realising a generational performance lift in this test.

Rainbow 6 Siege

Rainbow 6 Siege has an inbuilt benchmark which we’ve found very consistent.

Here the i9-11900K falls about 20 FPS behind the 10850K on average but is 60 FPS behind the 5800X. Obviously, all three CPUs develop very high performance but it’s a shock to see Intel’s latest flagship unable to outperform either their last generation of AMD’s current equivalent in this highly CPU dependent game.

Doom Eternal

Doom Eternal is also very well optimised and capable of high frame rates and we logged two minutes of play to give us these results:

The 11900K and 10850K perform nearly identically here again, with the 5800X clearly in the lead demonstrating that even with higher settings we’re not GPU-limited in these tests thanks to the power of the RTX 3080.

Shadow of the Tomb Raider

Moving on to more demanding titles, Shadow of the Tomb Raider’s inbuilt benchmark has exceptional consistency and gives us a breakdown of CPU performance, it’s those numbers we’re looking at here to completely isolate it from GPU performance.

This test is a close-run thing, the i9-10850K is marginally behind, the 5800X marginally in front on average. In reality, it’ll be your GPU that dictates performance in this game, but we’re seeing a trend in performance emerge now between these three CPUs.

Red Dead Redemption 2

Red Dead Redemption 2 hands another win to the Ryzen 5800X.

Again it’s surprising to see the newest CPU bringing up the rear here, 15 FPS on average behind the 5800X and slightly behind the 10850K.

Flight Simulator 2020

And finally, the game that places the biggest demand on CPU power here, Flight Simulator 2020. This benchmark comprises a three-minute flight from La Guardia over Manhattan and delivers a stern test of the CPU. GPU utilisation stays under 70% here and performance is ultimately dependent on CPU speed.

Here the i9-11900K outperforms the 10850K across the board, delivering 61 FPS on average. That’s not a bad score by any means, but the 5800X beats it once again at 63 FPS average, although performance is slightly less consistent with lower lows, 1% lows and 0.1% lows. Intel made bold claims in their launch presentation about the 11900K’s performance, stating that it was capable of beating the 5900X by 11% – it’s possible that that is the case in other tests or different circumstances, but in this benchmark, it’s not the case falling slightly behind on average.

Gaming performance conclusions

Rounding up the game testing sees the Intel Core i9-11900K in an interesting position: We’re used to seeing the latest component develop a commanding lead. In these tests Intel’s new flagship, the i9-11900K, not only failing to beat a six-month-old part from AMD, but on occasion struggling to match the last generation part from Intel themselves, and one that’s not even their top-flight product.

Memory Speed Scaling

There’s been some discussion online about memory ratios – ‘Gear 1 and Gear 2’ modes in relation to the i7-11700K and i9-11900K. It also helps explain how we arrived at our memory settings for these benchmarks. We’ll touch on this now to cover key points but if it interests you please see our companion article which digs deeper into the effects of memory latency on performance for this CPU and the i9-10850K.

Gear 1 and Gear 2 are simply the full speed or half speed memory controller ratios for the CPU to control RAM. Much like Ryzen’s ‘uclock’ setting this controller to half speed induces latency, and that latency induces a performance penalty. 

Let’s look at a couple of A-B tests in our most consistent benchmarks to demonstrate this effect: 

You can see that ‘Gear One’ offers a slight performance bump, a few FPS, but it’s not a marked difference.

RAM speed also has its own impact on latency. To demonstrate here’s a series of runs of Shadow of the Tomb Raiders benchmarks at different RAM frequencies, but timings retained at CL 16-16-16-32 up to 3600Mhz, and CL17 at 4000MHz for stability. We’re running Gear 2 throughout here because Gear 1 wasn’t stable at 4000MHz: Remember this CPU is only officially rated up to 3200MHz or a 1600MHz Memory Clock speed because the actual RAM clock speed is half the transfer speed.

You can see how the performance gain is significant, but peaks at around 3600MHz and tails off at 4000MHz because we have to loosen timings to maintain stability. The detriment of running 2400MHz RAM is serious, and this data challenges the notion that ram speed is unimportant to Intel CPUs or less important than Ryzen. It clearly makes a big difference to potential performance. This is why we felt it was vitally important to give this CPU the same advantage as the 5800X, and as it happens that occurs around the same RAM settings, 3600MHZ CL16 and gear 1. Overall RAM latency clearly has a big impact on this CPU’s performance. 

If you’d like to see a more in-depth analysis of this including data from the 10850K, please read our linked article focussing on the topic here.

Power and thermals

Power draw and the consequent heat output has long since been the cost of high performance on Intel’s 14nm CPUs. We ran tests to explore this on the i9-11900K. 

The most illuminating result was using the all-core load in Cinebench, and toggling Thermal Velocity Boost to ascertain its effects on both CPU temperature and power draw. These numbers are reported by HWinfo64, total package power and temperature, and in both cases with the 280MM AIO running at full speed.

The first run to the left shows behaviour with the thermal velocity boost enabled – you can see that stock power limits are enforced and the CPU regulates power to 250W. The ASUS motherboard allows this behaviour in its default configuration. All cores sit at about 4.7GHz and the CPU does a good job of holding temperatures at 70°C.  In the second run to the right, disabling Thermal Velocity boost actually allows the CPU to disobey power limits to achieve and maintain as high clock speeds as possible and it goes pretty wild, drawing up to 330W and hitting its new target of 90C before backing off the power and clocks to prevent overheating. Before that, a few cores are hitting 5.1GHz with most at 5Ghz. As a result of over-riding the power and thermal constraints, it scores 6042 points vs around 5900 points in the first run where the lower power limit is enforced. 

This second run is very much a ‘gloves off no limits’ approach, with normal behaviour overridden just to demonstrate the kind of power draw you may encounter if you’re looking to overclock this CPU. The first run is much more indicative of ‘normal’ behaviour and power draw, although in most cases after the higher power time limit, Tau expires, the package power will drop to 125W for extended full core loads. 

Another result of note is that simply changing the CPU cooler settings from automatic behaviour where it scales speed with CPU temperature to full speed all the time yields a 100 point increase in Cinebench R20 – cooling the CPU more aggressively and holding lower temperatures allows it to achieve higher performance. 

The power draw of this CPU can be pretty insane, and you do need both a very solid motherboard power delivery set up and a high-end cooling solution to get the best of it, particularly if you intend on overclocking it.

Conclusion

So, where does this information leave us?

This CPU is a disappointment. We’ve got Intel’s flagship product here, and yet we see it fail to consistently outperform their last-generation chip, and fail to beat the primary competitor from AMD.
Let’s not pretend Intel haven’t tried: They’re used to the top dog position and if they could beat AMD they would. The Zen 3 CPUs were released six months ago so there was a clear target to aim for, and in the synthetics, we can see that they’ve matched it, like for like. But in the gaming tests, it can’t compete.

Ultimately what we’re seeing here is the consequences of the limitations of that ageing 14nm process. At 10nm perhaps this CPU would have run cooler, more efficiently, and at a higher clock speed. Perhaps it would have less cache latency helping gaming performance. But that’s not the case. Intel has laid it all on the table and this is it. 

Then we come onto the real issue, which is price. This is a $539 product. You have to ask yourself what justifies that price? The raw performance doesn’t, and to cap it all you need to invest at least $250 more in a motherboard plus a top tier cooling solution to support it. Not only is the Ryzen 7 5800X $100 cheaper, but it also delivered the results here running on a $150 motherboard. For the cost difference, you could have a 5900X and 12 core performance that blows the 11900K into the weeds for any application that can make use of them. 

This 11^th CPU generation needs to be viewed as what it is – a stopgap that brings Intel up to the specification of Zen 3 chips, with native PCIe 4.0 support but cannot compete on raw performance. It’s also the end of the line for this process, a representation of what many years of refinement and tweaking can do but also what it can’t do. It can’t beat the competition. 

An area we haven’t assessed is the performance of the new integrated GPU – it has some features that may make a significant difference if you do a lot of video encoding or transcoding and again Intel make some bold claims in their productivity slides – so if you’re considering the 11^th Generation for a PC focussed on those tasks it will pay to dig out more specific benchmarks. 

Finally, Asus released yet another BIOS just 5 days before this release, giving us insufficient time to re-test and revalidate all our results. It claims to enable ‘Adaptive Boost Technology’ for this specific CPU, the only one in the product stack to use it. That may give a small bump in multi-core workloads in a correctly configured system, but given that it’s a Beta, and this CPU has actually existed for some time prior to launch, we don’t see it making a step-change in performance. It’s something we’ll review later.

Alternatives?

Ultimately, if you need a PCIe 4.0 platform for content creation or high-performance computing you’ll be looking at AMD anyway, the Ryzen 9 5900X and 5950X are seriously performant parts when available.
If you want a very powerful CPU on a budget then Intel caters to that at the moment with the i9-10850K which has been as low as $320, the i7-10700K or if you do want Rocket lake then I cannot see there is a huge gap in performance between this i9-11900K and the i7-11700K beneath it – it’s still an 8-core, 16-thread parts with very good gaming performance and more than enough versatility. And of course, if it’s just gaming you’re interested in, then the Zen 3 Ryzen 5 5600X and 7 5800X which are now more readily available offer the same or better performance at just $300 and $450 respectively, and with a lower platform cost, whilst the i5-10600K is discounted, doesn’t need as expensive of a motherboard, and offers excellent gaming performance as well. 

Nice try Intel, but sadly this CPU just isn’t good enough to justify its price tag. The box is really lovely though. 

The post Intel Core i9-11900K Review: Intel’s Last Stand | Performance Analysis vs 5800X vs 10850K appeared first on PremiumBuilds.

We’ve obtained examples of the RTX 3060 Ti, the RTX 3070, and the RTX 3080 and have spent the last two weeks running them through a suite of benchmark tests. In this article, we’ll present our results to you and conclude with some remarks as to what systems and usage cases these cards are best suited to. 

We’ve tested first-person shooters and AAA titles at 1080p, 1440p 1400p Ultrawide, and 4K. Our testing methodology differs from some other sites: We’ve chosen settings that show these games in their best light, and that’s not always ‘ultra’ across the board. We’ve tested most games here on high or a mix of high and ultra settings, allowing the GPUs to shine and giving you a better understanding of how you can expect them to perform in the real world. The settings remain consistent both within each test and across the resolutions.

From the results here, you will be able to tweak settings to obtain higher quality or higher framerates, whichever you prefer, but you can expect performance on a par with our numbers here. Where we identify a clear CPU bottleneck we’ve mentioned it (Hi Warzone!).

The test system is our Ryzen test bench with a Ryzen 7 5800X, 16GB of RAM running at 3600MHz Cl16, and a B550 motherboard.

So, let’s dig into the numbers. 

RTX 3060 Ti vs 3070 vs 3080: Synthetic Tests

3DMark

Firstly, a quick look at some synthetic benchmarks helps us verify our cards are performing correctly, and get an idea of where they stack up overall. We’re just looking at the GPU scores here. We’ve included some popular GPUs of the last generation so you can see how they compare.

Fire Strike uses DX11 which is an older API now, and we get an idea of how closely matched the GTX 1080 Ti and RX 5700XT are, and in this case the RTX 3060 Ti as well.

Time Spy uses The DX12 API and is more representative of current games, here the RTX 3060 Ti fares better but it’s clear the RTX 3080 still has a big lead.

Finally, we can take a looking ‘Port Royal’ helps us assess the relative Ray tracing capabilities of the RTX-enabled GPUs. This test uses DX12’s Ray tracing commands to render out a complex scene full of shadows, reflections, and lighting sources over and above the traditionally rasterised rendering of the core 3D scene. Despite using the last generation RTX cores the RTX 2080 Ti acquits itself well here, and there’s a fair gap between the performance of the 3060 Ti, RTX 3070, and RTX 3080 commensurate with their RTX core counts and overall rendering ability. The RTX 2060 brings up the rear, the least powerful card with RTX capabilities.

RTX 3060 Ti vs 3070 vs 3080: Gaming Benchmarks

1080p Performance

1080p FPS Gaming

Doom eternal is really well optimised and shows good scaling with GPU power, everything from the RX 5700XT up exceeding 240FPS.

In Call of Duty Warzone, on high settings, we see all the Ampere cards exceed 200FPS average but you’re not gaining a huge amount from the additional cost and power of the RTX 3080. I think we’re seeing the ceiling of the 5800X’s performance here although lows don’t suffer like the can in some titles when CPU bound. There was a hard limit of 250FPS maximum frame rate in all the Ampere GPUs.

Rainbow 6 Siege shows good 1080p scaling but even an entry-level GPU is capable of fairly insane framerates at 1080p in this game.

1080p AAA Gaming

As for the AAA Titles, In Shadow of the Tomb Raider we’re seeing all the GPUs exceed 144FPS average at 1080p with the 3080 nearing 200FPS. Of note here is that the 3070 was 78% GPU bound, and the 3080 was only 50% GPU bound: Even with a Ryzen 5800X, GPU performance is being left on the table in this title with the higher performing Ampere GPUs in this game at 1080p. 

In Red Dead Redemption 2, a more demanding title, the high settings load GPU’s even at 1080p. The 2080Ti and 3080 are the only GPUs to exceed 100FPS average, but in this title we’re really looking for a consistent 60FPS at high settings and everything from a 1080ti and 5700XT upwards is capable of that here. Like Doom eternal, the Vulkan API’s help the 5700XT perform – AMD cards respond well to this API.

1080p Flight Simulator

And finally in Flight Sim 2020, 1080p means the CPU performance is as exposed as GPU  performance, we see minor performance improvements from the higher tier cards but even the 3080 only just hits 60FPS: That’s a function of the demanding nature of this test, flying over New York City at low level. As you’d expect, at 1080p all of these GPUs perform very well and we do see utilisation drop below 100% on occasions as they wait for the CPU to process the game data. Ultra settings sees performance drop about 10FPS across the board.

1440p Performance

1440P FPS Shooters

This resolution is a much better match for these GPUs, and Doom Eternal is really well optimised letting the higher-end hardware shine. Again we see the power of the 3080 compared to the 2080 Ti, and Nvidias claims re the 3070 matching the 2080Ti ring true. Doom has a very well refined settings system: Each graphics preset adds or removes about 10FPS so you can fine-tune the game to your liking but all of these GPUs give a fluid, exciting experience at 1440p and over 100FPS, even the 1660 Super hits 90FPS average. It’s still very enjoyable but you might want to reduce settings a notch or two. 

Rainbow Six Siege shows excellent scaling with the more powerful GPUs, and the RTX 3080 really stretches its legs to deliver a 400FPS average: Even the 3060 Ti exceeds 260FPPS, and we’re well into diminishing returns since we’ve well exceeded even a fast monitors refresh rate and generally this is a title that would be played at 1080p. Note the 2060KO and below are run on a Ryzen 3600 System: That likely accounts for some of the performance difference here, but it’s not a huge difference.

Call of Duty: Warzone we still see them all exceed 144FPS average at high settings. The RTX 2080ti marginally outperforms the Ampere cards here, which could be down to it being an established card with optimised drivers whilst the Ampere cards are newer – but note they don’t scale well. An RTX 3080 isn’t getting you substantially more performance in this title over the much cheaper 3060 Ti.

1440P AAA Titles

Moving on to Triple-A titles, Red Dead Redemption can make the most powerful graphics cards sweat, and we have turned up settings here. That’s for a couple of reasons, firstly because the game should be experienced in all its glory at higher settings, and secondly, because it shows the significant break in performance between the cards that can maintain 60FPS average, and those that can’t. Last generation cards struggle, whilst the 3060 Ti comfortably maintains nearly 70FPS, and the RTX 3070, RTX 2080 Ti, and RTX 3080 all approach 100FPS average. Again, you can tweak settings to get the performance you’re happy with as they are turned right up here, but those GPUs are capable of providing an excellent experience at 1440p.

Looking at Shadow of the Tomb raider, again this title shows where the GTX 1660 Super begins to struggle, although it is still capable of running well optimised titles like this at 1440p. The scaling with the more powerful GPUs is clear with all of the Ampere GPUs exceeding 100FPS average. This is at ‘highest’ settings and we can see that the 3080 is the card that manages to exceed 144FPS average at 1440p in this title with settings cranked. 

In Flight Simulator at 1440p, we’re now becoming much more GPU bound. 50-60FPS average at high-end settings really is a good performance in this test, and it looks utterly gorgeous doing it. All Ampere GPUs exceed 30FPS at all times. 1440p is where Flight Sim 2020 really begins to shine, and the RTX 3060 Ti, in particular, impressed me with its performance here. Lifting settings to ‘Ultra’ adds some nice visual tweaks but still costs about 10FPS across the board. Again, there’s room to tweak here with just a couple of detail settings turned down from ultra getting you most of the performance benefits at almost no detriment to visuals. 

1440p Conclusion

We can see that these cards really begin to come alive at 1440p, significantly outpacing all but the highest-end options from the last generation. It was the 3060 Ti that impressed me most, with performance very much in the same ballpark as it’s bigger siblings. Once you’re up to 100FPS in AAA titles at high settings, you really are getting what you paid for in games, so there’s not much more to be asked for the 3060 Ti. The RTX 3070 also excels at this resolution. 

1440p Ultrawide Gaming

Moving on to ultrawide 1440p, this resolution bridges the gap to 4K and is becoming increasingly popular. We particularly like the versatility of workspace it offers, combined with the immersion in gaming.

1440p Ultrawide FPS Gaming

The first person shooters still record high FPS if they’re well optimised, Rainbow 6 Siege is generating over 250FPS in high settings. COD:Warzone really does seem to be CPU limited – FPS doesn’t really drop much from 1440p so you’re at no penalty opting for ultrawide. Doom Eternal is still generating well over 144FPS on everything except the 1080ti, and we’re at ultra nightmare settings here.

1440p Ultrawide AAA Titles

Moving on to AAA Titles, we’re exceeding 60FPS in Red Dead Redemption 2 on all the Ampere cards at demanding settings. The 3060ti and 3070 are again relatively Close in performance with 10FPS between them, with only the 3080 distinguishing itself 20FPS higher. Shadow of the Tomb Raider is much the same pattern, but with slightly higher framerates. The 3070, 3080, and 2080 Ti all approach or exceed 100FPS. 

And finally, at 1440p Ultrawide, we see the same scaling in Flight Simulator 2020 – But all of these GPUs are approaching or exceeding 50FPS even at this resolution which is impressive, and again you can see you’re paying a great deal in going from an RTX 3060 Ti to an RTX 3070 and on to an RTX 3080 to go from 46 to 60 FPS.

The takeaways here are really how closely the RTX 3070 matches the RTX 2080 Ti so Nvidias claims ring true there. Also, note that the hierarchy of cards is now well established but the 3060Ti really isn’t a bit-part player here: It’s very much keeping pace and in any of these titles you’re not going to see or feel a difference in performance between say 160 FPS and 190Fps, or in just a couple of settings turned down to see it match the performance of the RTX 3070. The 3070 is clearly excellent in 1440p ultrawide, but the 3060ti is more than capable as well. The 3080 holds it’s clear lead.

4K Performance

4K FPS Shooters

Moving on to 4K, we’re now at a resolution that can make these GPUs really work hard. Looking at the shooters, note that we’re still exceeding 100FPS in Doom Eternal and CoD Warzone – and we’re still at high settings here. Likewise, in Rainbow6 Siege we’re exceeding 144FPS meaning you’ll be maximising a high performance 4K monitor, not that this is an ideal set up for competitive play. The RTX 3080 is still pushing past 300FPS, but the 3060Ti and 3070 are much lower here beaten even by the GTX 1080ti and I confirmed this with multiple runs. I can’t confirm exactly why this is.

4K AAA Titles

AAA titles are where you can revel in the detail rendering at 4K gives – but we’re starting to see all of these GPUs struggling to make the magic 60FPS. You’re going to have to turn down settings a little from those we’ve used throughout these benchmarks. Again we see the 3070 shadowing the 2080 Ti closely although the 2080 Ti does come out on top – just – in these three tests. The 3080 has a clear advantage, and if you’re building a 4K focussed gaming PC then it’s really between the RTX 3080 and the RX 6800 XT as to which GPU is right for you.

Conclusions

The RTX 3060 Ti

Firstly, the RTX 3060 Ti was the standout card of this test. It performs excellently, to the point where it’s close to the RTX 3070 in most of these titles at 1440p. In well optimised games and first-person shooters, it’s producing fluid responsive gameplay at high FPS and high settings. In demanding titles, it’s punching well above its weight. The RTX 3060 Ti a great card, and if you’re looking to trade off components in your system to afford a better CPU, monitor, or SSD, you should absolutely consider the 3060 Ti. It won’t feel like a compromised choice at all.

Ampere Gaming at 1080p

Secondly, none of the Ampere cards achieve their potential at 1080p. They’re either exposing CPU limitation in very high FPS titles or else not shining in AAA titles restricted by a lower pixel count. You can make a case for the RTX 3060 Ti in either a very high FPS esports build, or in a PC aimed at playing AAA titles at 1080p and very high settings: But really, We’d recommend that if you’re considering any of these Qmpere cards you start your search with a 1440p Monitor capable of 144Hz and adaptive sync. That’s where you start getting your money’s worth out of these GPUs. 

The RTX 3080

Finally, of course, there’s the RTX 3080. This GPU is an absolute monster. It cleanly wins every benchmark here, and whilst the 3070 has clearly been pegged to the RTX 2080 Ti, and the RTX 3060 Ti massaged to ensure it’s 10% or so slower than that, the RTX 3080 is allowed to stretch its legs and hit the limits of its’ capability. That really comes down to power restrictions but We’ll dig into that more with a specific look at the RTX 3080 in a future article.

Watch out for bottlenecking in aged systems

One note of caution here is if you pair these GPUs with an aging system, they’re all-powerful enough to expose weakness, particularly in CPUs. You may find that whilst you can run at higher settings, you begin to experience stutter and frame time inconsistencies if you hit the limit of your CPUs performance. That’s something we want to investigate and we will be running through some tests with a Ryzen 5 3600 and Intel CPUs from the last-generation to see if they negatively impact performance at all, or if they’re sufficient to maximise the potential of these GPUs. 

Our Recommendations

Best Ampere Card for 1080p FPS Gaming / 1440p All-rounder Builds: RTX 3060 Ti

Based on our results, realistically we’d recommend the RTX 3060 Ti as the high FPS esports GPU at 1080p, or for a 1440p all-round gaming machine.

Related: RTX 3060 Ti Aftermarket Card Overview
Related: RTX 3060 Ti Aftermarket Card Database

Best Ampere Card for 1440p AAA Gaming: RTX 3070 / 3060 Ti

Stepping up in to solid 1440p AAA title performance the RTX 3070 makes a strong case for itself, but cross-shop the RTX 3060 Ti if you’re working to a budget – you won’t feel short-changed with its performance and if it allows you to purchase a high-performance CPU as well then it’s the right choice to make.

Best Ampere Card for 1440p Ultrawide Gaming: RTX 3070 / 3080

At 1440p Ultrawide, the RTX 3070 shines but the additional cost of a RTX 3080 begins to make sense to provide a really remarkable gaming experience. 

Related: Best RTX 3070 Aftermarket Cards
Related: RTX 3070 Aftermarket Card Overview
Related: RTX 3070 Aftermarket Card Database

Best Ampere Card for 4K Gaming: RTX 3080

And finally, at 4K, it’s such a demanding resolution that in high-end titles like Red Dead Redemption or Cyberpunk you will need to lower settings from Ultra to maintain 60FPS performance even on an RTX 3080. There is one ace up the Ampere GPUs sleeve though: DLSS. This technology form Nvidia uses a trained neural network and AI to up-sample the rendered image for display: when it works well it allows you to combine heightened settings with higher frame rates. However, just a handful of titles are using it to it’s full potential, so it’s something we’re looking into and will deliver our impressions on it later.. 

Related: Best RTX 3080 Aftermarket Cards
Related: RTX 3080 Aftermarket Card Overview

I hope you’ve found this roundup of the Ampere GPUs useful: pair any one of these cards with a well-matched monitor (linked below) and you’ll have an absolutely fantastic set up to enjoy the latest games as they are meant to be played.

• Best Monitors for RTX 3060 Ti
• Best Monitors for RTX 3070
• Best Monitors for RTX 3080

The post RTX 3060 Ti vs 3070 vs 3080: Benchmark Comparison (Real World Tests) appeared first on PremiumBuilds.

Nvidia’s Ampere launch has brought a large step forwards in performance at every price point. The three mainstream cards currently released, the RTX 3060 Ti, RTX 3070 and RTX 3080 all excel in high resolutions and the most demanding current games. The RTX 3060 Ti and RTX 3070 are closest in specification and price, with the RTX 3060 Ti priced at around $400 and the RTX 3070 starting at $500 for the various partner cards. In this review, we’ll take an in-depth look at those two cards to find out which is going to be best for your needs. We’ve run hundreds of benchmarks and spent many hours tweaking these GPUs to best inform you of their relative strengths, so let’s dig in!

RTX 3060 Ti vs RTX 3070

Specifications

	RTX 3060 Ti	RTX 3070	RTX 3080
Design
Price (MSRP)	$399	$499	$699
VRAM	8GB GDDR6	8GB GDDR6	10GB GDDR6X
VRAM Bus	256bit	256bit	320bit
GPU Core	GA104-200-A1	GA104-300-A1	GA102-200-KD-A1
Process	Samsung 8N	Samsung 8N	Samsung 8N
Cuda Cores	4864	5888	10496
RT Cores	38	46	68
Tensor Cores	152	184	272
PCie Bus	PCIe 4.0x16	PCIe 4.0x16	PCIe 4.0x16
Pixel Rate	133.2 GPixel/s	165.6 GPixel/s	164.2 GPixel/s
Texture Rate	253.1 GTexel/s	317.4 GTexel/s	465.1 GTexel/s
Base Clock	1410 MHz	1500 MHz	1440 MHz
Boost Clock	1665 MHz	1725 MHz	1710 MHz
TDP	200W	220W	320W
AIB Cards	RTX 3060 Ti Card Database	RTX 3070 Card Database	RTX 3080 Card Database

The main similarity between these cards is in their memory specification: Both have 8GB of GDDR6 Memory operating at 448GB/second bandwidth, across a 256bit bus. They forgo the faster GDDR6X memory specification of the RTX 3080. 

The differences lie in the core. Whilst both use the GA104 core, the RTX 3060ti makes do with a cut down version of it. The RTX 3070 has 5888 shader units, 184 tensor cores, and 46 Ray tracing cores. It boosts to around 1725MHz. The RTX 3060Ti has 1000 fewer shader units at 4864, 152 Tensor cores, and 38 Ray tracing cores. It also boosts slightly lower in reference specification, to 1665MHz. In operation, both comfortably exceed their reference boost clocks.

In real terms, we should see performance commensurate to those key specification differences. 

To test this, we’ve got our hands on an example of each card. The 3060 Ti is a Zotac Twin Edge OC, a twin fan 2 slot model. This card claims a 1695 MHz boost clock but our example self boosts to sustain 1860MHz and is unusual in that it’s a pure 2 slot design and relatively short at just 222mm. Other than that it has the same specifications as any other RTX 3060Ti.

The RTX 3070 is EVGA’s XC3 Ultra model. It’s a 3 fan card with a ‘2.2’ slot thick cooler, and EVGA claims a boost clock of 1770Mhz for this model, but in fact, it itself boosts to 1950 MHz happily and will sustain a 2150Mhz boost when overclocked. 

All tests were conducted on an identical test rig: The System comprises of a Ryzen 7 5800X CPU with PBO enabled, A MSI Mortar B550 motherboard, and 16GB of RAM clocked at 3600MHz Cl16 and operating in dual channel mode. Power Supply is a 650W Gold Antec unit.

‘Real World’ Testing

A note on how we test: These tests are conducted on what we consider ‘real world’ settings. That is, High or Highest unless the very highest settings are a serious detriment to performance. We’re aiming to present to you the cards performance as close as possible to how you will experience them on your own PC, whilst playing the games you love. The tests are completely standardised and identical in each benchmark so the cards are being tested like-for-like. Please see the footnote for details of each benchmark.

Performance

Firstly let’s check out some synthetic benchmarks, to see how the GPUs perform in a standardised test and also to verify that our examples are comparable to other versions. We’ll also take a quick look at overclocked performance just to see how much performance there is on the table, although this isn’t an in-depth look at overclocking these cards. We’ve included the GTX 1080 Ti and the RTX 2080 Ti for reference.

In Firestrike we can see the clear segmentation between these cards – the stock 3060 Ti matches the 1080 Ti in this Direct X 11 test that renders at 1080p. The RTX 3070 makes good its promise of matching the RTX 2080 Ti, whilst the RTX 3080 sits in a league of its own at the top. You can see an overclock adds a little to the RTX 3070 and 3060 Ti, but doesn’t bridge the gap between them and the next card up.

In Time Spy which is direct X12 and renders at 1440p, the Ampere cards come into their own with the 3060Ti clearly beating the 1080ti. The RTX 3070 perfectly brackets the 2080 Ti with it’s stock and overclocked configuration, whilst again the RTX 3080 sits at the top over 2000 points clear. 

To evaluate Ray Tracing performance we can look at the Port Royal Benchmark, which yields a single score based on overall GPU performance. Here again, we see the RTX 3080’s clear lead, but focussing on the RTX 3060ti we can see it is 1000 points short of the 3070, and an overclock does little to help. The 3070 gains a little more from an overclock, but still falls a little way short of the claimed ‘RTX 2080ti beating claim’.

How does this translate into actual gaming performance though? We’ll focus on the two cards in question here to work out the differences between them.

Gaming Benchmarks

Looking first at some fast-paced shooters, COD Warzone shows how capable both of these GPUs are at 1080p and 1440p, and even at 1440p ultrawide, we’re still at around 144FPS making full use of a high refresh rate monitor. We’re using high settings here so you’re not even sacrificing fidelity for speed and you could push framerates higher with lower settings. This game is actually heavily dependent on CPU as well so if it’s the game you care about, you’ll get better results with the 3060Ti and a great CPU than you will with the 3070 and a compromised CPU. By 4K we are struggling to exceed 100FPS so I wouldn’t recommend these GPUs – or playing warzone in 4K on them. Note how close the two GPUs remain across the board with just 10-15FPS separating them.

Rainbow 6 Siege is even faster-paced, and again both GPUs exceed 240FPS at 1080p, 1440p and 1440p ultrawide. They even exceed 144FPS at 4k, and this in on very high settings across the board. Again, we see limited performance improvement from the additional spend on an RTX 3070, and CPU performance is always going to be more important here.

Finally, as an example of a well optimised first-person shooter, Doom eternal on Ultra Nightmare settings shows strong scaling both with hardware and resolution. Again, we see very high framerates of around 300FPs average on 1080p, over 200 for both the 3060Ti and 3070 at 1440p, and well over 144Fps at 1440p ultrawide. Here, about 20FPS separate the 3060Ti and 3070 with settings like for like, but realistically the difference to the player of 205 FPS vs 225 is minimal, both provide an excellent experience and exceed the refresh rates of very high-performance monitors at 1440p and 1440p ultrawide, and it does well at 4K too.

Moving on to more demanding AAA titles, this is where we perhaps see a little more separation between the two cards.

Red Dead Redemption’s inbuilt benchmark gives us an insight into the demands of current games. These have all been run at identical settings, a mix of high and ‘ultra’ and favoring visual quality to best show off the game’s graphics. In reality, you can drop settings for more performance at no real cost to the visual quality of the game. Nonetheless exceeding 60FPS is our target here. Even at 1080p, we see averages below 100FPS: That’s a function of both of the demands of the game but also the fact that the rendering pipelines in Ampere are set up to be more efficient at higher resolutions so you’re not seeing their full potential at 1080p. Moving on to 1440p both GPUs exceed 60 FPS, but the 3070 scores 87FPS to the 3060Ti’s 69 FPS. Both are totally playable, but you can see that you will have to lower settings a bit more on the 3060ti for equivalent performance – manually forcing settings to ‘high’ except keeping textures at ultra sees performance to 81FPS on the 3060Ti, but keeps the games visual flair intact. At 1440p ultrawide again we see that both remain playable but there’s a hit to the performance with 59 FPS on the 3060Ti and 68 on the 3070. Again, moving to high in place of ultra on the 3060Ti equals the 3070’s performance here and both GPUs remain very enjoyable to play.  At 4K, still on ultra, we’re seeing 45 and 50 FPS respectively and this will probably look jarring to many people. We need a mix of medium and high settings on the RTX 3060 Ti, and high and ultra on the RTX 3070, to hit 60FPS average. Getting the very best out of titles like Red Dead 2 at 4K needs more GPU power but if you’re willing to accept some compromise to settings either card plays acceptably well.

Shadow of the Tomb Raider is either less ambitious or just better optimised depending on your point of view, but still delivers state of the art visuals. At the highest settings, you’re getting the full effect and at 1080p you can see both GPUs comfortably exceed 144FPS. At 1440p both are just above 100FPS, with the 3070 around 15 FPS ahead, but this is an excellent result for both. Moving up to ultrawide, we’re at 81 and 94 FPS respectively, again close enough to not see a vast difference when gaming, and within reach of a few settings tweaks to equal performance.  At 4K we fall under the 60FPS target with the 3060 Ti struggling a little at 54FPS, but the 3070 managing 63 FPS average. The game will play acceptably well at 4K on these cards, and this is at the very highest settings. 

And finally, Flight Simulator 2020 requires a brutal mix of next-generation graphical and CPU power, and you can see that by the fact we’re CPU limited at 1080p high settings on both cards with GPU utilization rarely exceeding 90%. This test comprises a 3-minute low-level flight over Manhattan and 60FPS is the highest performance possible on this system – not even an RTX 3080 exceeds it. At 1440p and 1440p ultrawide, the 3070 takes a 10% lead over the 3060 Ti and extends it to more like 20% at 4K. Both of these GPUs are exceptionally capable in this title at 1440p and 1440p ultrawide, and you can mix ultra and high settings for stunning scenery and near 60FPS performance. As the first-person shooters, this is a title that actually needs careful consideration of the whole system to get the best performance out of it. See our Flight Simulator Performance Analysis guide for more benchmarks and recommendations for this title.

Cyberpunk 2077

Much of the hype around the Ampere cards has focussed on Cyberpunk 2077, thanks to it’s highly demanding graphics and implementations of both RTX and DLSS 2.0. Whilst we have spent significant time playing it and testing it, we don’t feel it’s ready for anything more than a qualitative assessment at the moment. Performance is changing all the time with fixes and updates – hopefully for the better. Similarly, the complexities of performance variations with RTX on or off, and the various DLSS settings by their nature make it largely a matter of opinion as to what you prefer. DLSS is also somewhat of a ‘black box’ – AI-powered upscaling is by its nature hidden from the user and provides remarkable results but ones that may feel a little strange to players not accustomed to it. We’re planning a more in-depth look at these features soon, but for now, let’s consider it like this: Do these GPUs allow you to play and enjoy the game, and does the RTX 3070 offer a significant advantage over the RTX 3060Ti?

Both GPUs allow 60+FPS  in-game without using RTX or DLSS at 1440p. Using RTX for lighting and reflections means you need DLSS turned on to achieve acceptable performance on either GPU, but even the 3060Ti manages 60FS with DLSS in balanced mode rendering at an internal resolution of 960p and upscaling to 1440p. It looks remarkably good doing this and it’s likely going to come down to personal preference how you want to play the game. The RTX 3070 is powerful enough to also run RTX shadows and higher quality RTX lighting and maintain 60FPS, or simply hit higher frame rates. There’s so much customisation built into just the RTX and DLSS settings that you’re bound to find a compromise that works for you and this goes for either GPU. At 1440p ultrawide, both cards begin to struggle to hit 60FPS, and at 4K you are either looking at maximum assistance from DLSS or unacceptable performance, particularly on the 3060Ti. RTX isn’t viable with frame rates dropping to the 30’s. With RTX off and DLSS set to balanced the 306 0Ti is still capable of 60FPS at 4K though, showing the value of that upscaling technology.

For Cyberpunk 2077, either the RTX 3060 Ti or 3070 provide a more than acceptable experience at 1440p.

At 1440p ultrawide or 4K, we’d recommend the RTX 3080 to experience the full gamut of next-gen graphics this title has to offer. 

Value

Value is of course relative: If you want the higher performance of the RTX 3070 and are prepared to pay for it, then that represents value to you. However, looking at the various performance differences, we see the RTX 3070 is around 10% faster at 1080p, 12-14% faster at 1440p and ultrawide, and 15% faster on average at 4k. However it’s 20% more expensive at suggested retail price. Further, as we’ve shown in these tests the RTX 3060 Ti is just as capable as the RTX 3070 and it’s often just a case of slight settings tweaks to bring the cheaper card up to the same level of performance. Factors like the identical VRAM specification make it even harder to separate them.

If you want the best value then that’s clearly the RTX 3060 Ti.

If you’re prepared to pay for that slightly richer visual experience or extra performance, then the RTX 3070 brings that to the table for around $100 more. 

Conclusions and Recommendations

Best for 1080p High-FPS Gaming

At 1080p you’re not getting the best of these cards – although they’re obviously excellent performers. You’ll want to pair them with a top flight CPU to get as much performance as possible, and a very high refresh rate 1080p monitor. There’s just a 10% average difference here across all the titles, and we’re already achieving high framerates, so if you must go overkill at 1080p the RTX 3060 Ti is the right choice. Nvidia has a new GPU model release imminently and this should better cater for 1080p gaming.

Best for 1440p Gaming

At 1440p again the RTX 3060 Ti is a strong performer, with the RTX 3070 offering slightly higher framerates or visual settings, whichever is your preference. But there’s no night-and-day difference between these cards and if you’re working to a budget the RTX 3060 Ti is the right choice.

Best for 1440p Ultrawide Gaming

At 1440p ultrawide, you do see slightly more performance from the RTX 3070, and if you’re intending on playing AAA titles at that resolution it’s a great starting point. For occasional or more casual gaming the RTX 3060 Ti also does well.

Best for 4K Gaming

And finally, at 4K, either of these cards can provide an acceptable experience but if you’re planning a 4K Gaming PC then that really is still the realm of the RTX 3080. 

Footnote: Game settings used in benchmarks

The post RTX 3060 Ti Vs RTX 3070 Benchmark Comparison: Which is Best for You? appeared first on PremiumBuilds.

We’ve had a few days with our Zen 3 Ryzen 5800X and the bulk of that time has been spent investigating the impact of RAM speed on this new architecture. Since its inception, the Zen design has shown impressive performance scaling with RAM Speed. This a function of the design of the CPU itself.

In short, the CPU cores are connected to the rest of the PC via an interconnect chip known as ‘infinity fabric’. The speed this chip is running at is instrumental to the overall performance of the CPU. Luckily, AMD has an open attitude towards allowing users to tweak settings to their heart’s content, so all of this is adjustable in BIOS. The optimum settings involve ensuring that the RAM speed matches the Infinity fabric frequency 1:1. With these two clock speeds in sync, memory access latency is reduced and you enjoy a performance boost.

Previous Zen 2 RAM Investigation Summary – will Zen 3 continue the trend?

In our previous Zen 2 Ram Speed article we demonstrated the clear performance benefits of increased RAM speed with optimum performance at 3600 and 3733Mhz, and infinity fabric at 1800 and 1866Mhz respectively (note that DDR ram being Double Data Rate, it reports speeds at twice the actual clock frequency). Once you exceeded the ability of the Infinity fabric to match RAM speeds 1:1 however things broke down with increased latency costing performance despite the improved speeds.

The benchmark tests that showed clearest scaling were 3DMark Timespy, and Shadow of the Tomb Raider. Rainbow 6 Siege’s FPS also showed strong correlation with RAM speed. 

RAM Speed Scaling on a Zen 2 CPU, the Ryzen 3600. 

In this investigation we focussed on the following questions: 

• Firstly, does Zen 3 improve our ability to lift Infinity Fabric speeds to match faster RAM?
• Secondly, does Zen 3 performance scale with RAM speed in the same way we’ve become accustomed to?
• Finally, can we use this information to make any recommendations about the best RAM for your Zen 3 system?

Overclocking RAM on Zen 3

Overclocking on the Zen 3 platform proved relatively straight forward. Using our Patriot Viper 4400Mhz memory kit, which comprises Samsung B-Die Ram with CL19 timings, we were quickly able to establish that our Ryzen 5800X was stable at 4000MHz with Infinity Fabric (Fclck) at 2000MHz. 4066MHz was not stable. We were able to tighten timings down to CL 16-16-16-34 at 4000MHz. From there we established representative RAM profiles across a spectrum of speeds to emulate widely available RAM kits. 

Zen 3 RAM Synthetic Test results: CineBench R20 and Timespy

Honing in on the benchmarks that matter we again showed that Cinebench is not a reliable indicator of performance linked to RAM: Our scores varied little with RAM speed. 

Note that the scale starts at 5000 Points: The difference between these results is small enough not to cause concern, and it’s not possible to consider this a ‘trend’ in performance owing to the marginally lower scores at higher speeds. Cinebench R20 appears almost oblivious to the RAM speed of the System with even 2133MHz not causing a significant performance deficit.

TImespy is a Dx12 based benchmark with a portion dedicated to CPU performance in isolation. It’s those scores we’re concentrating on here.

In Zen 2 we saw a strong positive trend with RAM speed from 2133MHz to 3733MHz, and we observe the same here on Zen 3: Slower RAM is a serious detriment to performance, with 3600MHz yielding much higher results. However, increasing speeds into the realms of 3800 and 4000MHz sees little if any improvement in performance with 4000MHz CL18 close but slightly behind 3600MHz and 3800MHz. Tightening timings to CL16 sees performance slacken to just under 12000 points: A very interesting results given what we thought we knew about infinity fabrics speed and its relationship to CPU performance. We don’t know the reason for this given that all things being equal these tighter timings should increase performance, or at the very least not harm it. 

Moving onto key gaming benchmarks, we looked at Rainbow 6 Siege and Shadow of the Tomb Raider. Both of these titles demonstrated the benefits of faster RAM on Zen2 and again showed the negative impact once the 1:1 infinity fabric to memory clock ratio was broken.

Zen 3 RAM Gaming Benchmarks: RAM Speed vs FPS

1. Rainbow 6 Siege

Here we again see the same performance scaling to 3800MHz, followed by a slight tailing off as RAM speed climbs higher. Tightening timings at 4000MHz helps a little. Despite maintaining a 1:1 infinity fabric ratio throughout, we’re seeing a similar but less pronounced softening of performance past 3800Mhz just as we did on Zen 2.

2. Shadow of the Tomb Raider

This game again uses DX12 and has a comprehensive benchmarking tool which gives us interesting insight into the factors affecting the game’s performance. Specifically, it breaks down performance into ‘CPU game’ and ‘CPU render’ as well as GPU performance and indicates how GPU limited the benchmark is. All tests were conducted at 1080p medium on a GTX 1080 Ti but even that induced a GPU limitation due to the prodigious speed of the Ryzen 5800X. Due to the benchmark independently reporting CPU frame rates it can be eliminated as a factor.

We once again see a similar trend here with peak performance at 3800Mhz, and then a slackening at 3933Mhz and 4000MHz. From 3600-4000Mhz performance is in the same ballpark with variances that aren’t going to be material in the real world. Note that none of these will impact gaming at normal settings: You will be GPU limited at FPS below these numbers in this detailed AAA title. 

Conclusion: 3600MHz is still the best RAM for AMD’s Ryzen CPUs

We can now set about answering the questions we posed at the outset:

Zen 3 does allow us significantly higher Infinity Fabric clocks, and with it higher viable memory overclocks. This is exciting for overclockers because of the potential performance gains it brings.

However moving on to the second question, we do not see straight forward linear scaling until the break down of infinity Fabric Clocks as we did with Zen 2: In this instance, performance tails off even with a 1:1 infinity fabric ratio despite faster ram with tighter timings and lower latencies. In the real world, these slight variances won’t be noticeable, but similarly, we are not seeing anything like the scaling from 3600MHz-4000MHz as we get from raising Ram speeds from 3200MHz – 3600MHz.

Finally, we can now make confident recommendations on the basis of these results, which are in line with those obtained by experienced overclockers: 3600MHz RAM is still the best option for Ryzen CPUs. To arrive at this conclusion we can consider firstly that performance plateaus or even tails off after that. Secondly, there is the cost and availability of RAM. Since 3600MHz became the ‘sweet spot’ for Zen 2 Manufacturers have produced a number of fantastic RAM kits and prices have dropped due to a glut of RAM supply. To obtain RAM significantly faster than 3600MHZ prices jump from around $80/16GB for 3600MHz CL16, to $100 or more for 4000MHZ CL18 which as our investigation above shows will be marginally slower on Zen 3. Finally, investigations by Gamers Nexus indicate that dual-rank DIMMs may be slightly faster, making a 32GB kit in 2x16GB Configuration at 3600MHz likely to be ideal for most people.

With all of this in mind, we make the following recommendations for RAM for Zen 3 builds in 2021:

RAM Recommendations for Zen 3 Builds

Award	Design	Model
Best All-round RAM Kit for Zen 3 Builds		G.Skill Ripjaws V 32GB DDR4-3600 CL16
Best RAM for Zen 3 Gaming Builds		Crucial Ballistix 16GB DDR4-3600 CL16
Best Enthusiast RAM Kit for Ryzen Zen 3		G.Skill Ripjaws V 16GB DDR4-4000 CL16
Best RGB RAM for Ryzen Zen 3		G.Skill Trident Z Neo DDR4-3600 CL18
Best 64 GB RAM Kit for Ryzen Zen 3		Corsair Vengeance LPX 2x32GB DDR4-3200 CL16

Best All-round RAM Kit for Zen 3

Combining all that we have learnt so far about Zen 3 Memory performance, the G.Skill Ripjaws V 32GB DDR4-3600 CL16 kit hits all the bases. Using 2 dual-rank 16Gb Sticks with 3600MHz frequency and tight 16-19-19-39 timings mean you’ll be as close to optimized as one click can get you just by enabling the A-XMP profile. 32GB RAM provides ample for productivity, the most demanding games and light creative work such as video editing. At $150 it’s reasonably priced for 32Gb, exactly double the cost of the 16Gb equivalent. This is the RAM kit we’re confident meets the needs of almost everyone looking to take advantage of the prodigious power of the Zen 3 CPUs for gaming and general use. 

Best RAM for Zen 3 Gaming Builds

Crucial is Microns commercial brand and the Ballistix 16GB DDR4-3600 CL16 kit uses Micron E-Die RAM chips to deliver the performance sweet spot at a very attractive price. 16GB is still ample for almost all games and general use and multitasking. It’s been available for some time at the $75 price point making it a very worthwhile purchase and the best value of all the popular 16GB kits. The 3600MHz speed enjoys widespread compatibility with Zen 3 CPUs, whilst the quality of the E-Die chips lend themselves to tweaking and overclocking for even more performance if the mood takes you. If not the embedded A-XMP profile will ensure reliability and stability at the rated speeds. It is a 2x8GB stick for dual channel operation and any RAM kit you consider should be a dual-stick kit because of the large performance benefit that brings. The kit ships with discreet black anodized heat spreaders that integrates well with most builds but it is also available in white or red if that better suits your preferences. With this kit just $10 more than the most basic of RAM kits from less well-established brands we see no reason to opt for anything else. 

Best Enthusiast RAM Kit for Ryzen Zen 3

If you want to explore the limits of memory overclocking on Zen 3 then there’s no substitute for the Samsung B-Die RAM that makes up this kit, the G.Skill Ripjaws V 16GB DDR4-4000 CL16 – our recommended RAM kit for enthusiast Zen 3 builds. It’s very fast RAM with tight timings at the speed, and whilst it may not work optimally with the A-XMP profile applied there’s huge scope for manually tweaking timings and exploring optimizations. B-die kits can range from $100 to $500 or more depending on binning but it’s certainly not worth getting a top tier binned kit unless you’re intent on challenging world OC records. If you want to find out for yourself exactly what makes Zen 3 run sweetest, this is the kit for you. 

Best RGB RAM for Ryzen Zen 3

We’ve got a soft spot for a set of four RAM sticks with synchronized RGB here at Premiumbuilds and the G.Skill Trident Z Neo DDR4-3600 CL18 kit certainly scratches that itch. The core specifications are solid with 3600MHz and CAS latency of 18 so your CPU will operate with a near optimum 1800MHz infinity Fabric Clock. G.Skill offer a very useful reverse QVL list so it’s easy to ensure compatibility with your chosen motherboard. The RGB can be controlled via any of Gigabytes RGB Fusion, Asus Aura Sync, MSi Mystic Light or Asrock Polychrome software so it integrates well with the rest of your RGB and doesn’t require stand alone software. The RGB elements are beautifully detailed and this kit will lend some real visual flair to your build. 

Best 64 GB RAM Kit for Ryzen Zen 3

If your primary concern is the quantity of RAM – be it for a video edit workstation or scientific computation – then the Corsair Vengeance LPX 2x32GB DDR4-3200 CL16 kit is an excellent choice. Our benchmarks show minimal performance penalties at 3200MHz – RAM latency simply doesn’t impact tasks like rendering in the same way as it does gaming – but this is still 3200MHz RAM with CAS latency of 16 and performs well enough for AMD to showcase their new top tier GPU’s using RAM of this specification. Corsair are a long-standing brand and their LPX line is ubiquitous so obtaining matched RAM for future expansion won’t be a problem. Corsair also offer a 3600MHz LPX kit but supply is patchy and with a CAS latency of 18 the actual memory latency is nearly identical to this kit. If you need to pack as much RAM as possible then sacrificing a little speed for cost and capacity makes sense which is why this kit is a great buy. 

Benchmark Notes

The tests were conducted using the Premiumbuilds test bench, which comprises a Ryzen 5800X, MSI B550 Mortar Motherboard using AGESA 1.1.0.0 Patch C BIOS. RAM used was the Patriot Viper PVS416G440C9K 4400Mhz Samsung B-die Kit. PBO was activated but no manual CPU overclocking was undertaken. Timings under test were as follows

The post Zen 3 RAM Speeds: An Analysis of the Best RAM for Zen 3 appeared first on PremiumBuilds.

In this article, we’re comparing the two ‘real world’ flagships of the AMD and Nvidia GPU line ups: The RTX 3080 and the RX 6800 XT.

It’s only right that we start with an explanation both of the delay to this review, but also around the prevailing situation with GPUs. 

We’re not party to normal press channels to obtain cards so we’ve been subject to all the same pressures as the general market. We’re therefore only too painfully aware of the scarcity of GPUs across the board right now, and it’s part of the reason for the delay to this comparison article.

Of course, the global pandemic is one of the driving factors behind both restricted supply and also heightened demand for GPUs. On the one hand, production lines have been hit by a supply shortage and global supply chains have been disrupted. On the other hand, demand has skyrocketed with more people working from home or looking for means to entertain themselves. AMD in particular is struggling to produce sufficient GPUs on the 7nm production lines alongside the console APUs and Zen 3 CPUs that all vie for the limited capacity at TSMCs’ foundries. And finally of course Cryptocurrency mining has again come to the fore with miners buying up ‘gaming’ GPU’s in bulk to mine Ethereum. We all know the results of this: The cards we’re reviewing here sell via side channels for two to three times their launch MSRP, and are near impossible to buy via normal retail channels.

We’re going to keep this review straight down the line: You might be looking for one or other card and be prepared to pay. You might want to know if you can reasonably substitute one for the other and not suffer any major performance loss. This review simply seeks to answer the question: ‘how do these cards perform’, because it’s impossible to answer the question ‘are these cards worth your money’ until such time as the market stabilizes and they become available anywhere close to their suggested retail pricing.

First, let’s remind ourselves of the key specifications of these cards:

RX 6800 XT vs RTX 3080: Specifications

	AMD RX 6800 XT	Nvidia RTX 3080
Design
GPU Architecture	AMD RDNA 2	Nvidia Ampere
Process	TSMC 7nm	Samsung ‘8nm’
Transistors	26,800 million	28,300 million
VRAM	16Gb GDDR6	10GB GDDR6X
Memory Bus	256Bit 512GB/s	320bit 760 GB/s
Ray Tracing Cores	72 – RDNA2	68 – Ampere ‘2nd Gen RTX’
Tensor Cores		272
‘Enhanced Memory Access’	Smart Access Memory	Resizeable BAR
PCIe Version	4	4
Power Draw (tested)	300W	320W (340W)
Availability	Amazon.com	Amazon.com

It’s not possible to make direct comparisons between clock speeds or the number of shader units for these two GPUs – their architectural differences make the comparison meaningless.
We can directly compare memory specification and the RX 6800 XT has a massive 16GB VRAM, but it is just standard GDDR6. The RTX 3080 has 10GB of VRAM but it’s higher performance GDDR6X, produced by Micron under an exclusive partnership. This combined with a 320bit bus gives the Nvidia card higher memory performance but less capacity. In respect of additional features, Nvidia is obviously a step ahead in hardware ray tracing, but the RX 6800XT does also have hardware ray tracing cores and few games currently utilize this feature. Both have a mechanism to enhance memory access and boost performance – Smart Access Memory for AMD and ‘Resizeable BAR’ for Nvidia, but again this is specific to certain titles, and in the case of the RX 6800XT it can be detrimental to performance in some games. Finally, both have similar power draw requirements although the RTX 3080 is slightly more demanding, hitting 340W under peak load or when overclocked. 

The test set-up:

Both GPUs were tested in the same system – our Ryzen Test bench which comprises a Ryzen 7 5800X CPU, with PBO enabled but no additional overclocking. It runs 16GB of 3600MHz CL16 RAM and has a 1Tb Sabrent Rocket PCIe 4.0 NVMe SSD. It is powered by a Fractal Design Ion 860W Platinum power supply to ensure sufficient power. SAM and ReBAR were not enabled for any testing.

Synthetic benchmarks: 3D Mark

Running through the 3D Mark suite of benchmarks, we can see that in both of the straight ‘gaming relevant’ tests the RX 6800XT has the upper hand. 

In Fire Strike which is a 1080p rendered DirectX11 based test, the RX 6800XT turns in an absolutely stellar performance with a score of nearly 53,000 – that’s over 10,000 points or 20% ahead of the RTX 3080, an impressive achievement indicating very high performance in titles using DX11 and running at 1080p.

Time Spy, which uses the more recent DirectX12 API and renders at 1440p also shows a slight advantage to the RX 6800XT although the margin is reduced here to 1000 point or about 5%. 

Port Royal is a new benchmark that uses DirextX12 Ultimate to bring the GPUs hardware ray tracing to bear, and both cards run this benchmark fine but the RTX 3080 with its’ second-generation RTX cores has clearly better performance, by about 1,800 points or 20%. I’ve included the score of the RTX 2080ti here as well for reference, showing how closely the RX 6800XT matches the performance of Nvidia’s last-gen Ray tracing tech – although this score is an aggregate of rasterized and ray tracing performance and without a score break down it’s impossible to ascertain how much of that is the core rendering performance of the RX 6800XT and how much is the ray-tracing component. Nonetheless, it’s not fair to say the RX 6800XT is ‘poor’ at ray tracing, simply that it can only use DirectX12 Ultimate, and that it’s not at Amperes performance levels.

Blender

As a quick test of the relative rendering performance, we ran Blender 2.83s benchmark renders, BMW27 and classroom. Blender has a range of rendering options depending on the GPU used, and we tested cards in all compatible modes to get an idea of how they perform. The RTX 3080 can use CUDA or ‘OptiX’ acceleration, whilst the RX 6800XT can only use Open CL. The results aren’t favorable for the RX 6800 XT: The BMW render takes it 39 seconds to the 3080’s 11 seconds, whilst the classroom scene takes 74 seconds to the RTX 3080’s 41. The OPTIX rendering pipeline uses tensor cores to accelerate the process, and it’s dramatically faster. Even an RTX 2060 KO matches the RX 6800 XT in BMW27 using the OPTIX renderer vs Open CL, although the RX 6800XT is much faster at the Classroom render. This highlights how specific these benchmarks can be to one or other aspects of a GPUs performance.

Overall, if you’re intending on using these cards for productivity or 3D rendering, it does pay to look closely at performance in metrics specific to your needs and software. It may be a more readily available or cheaper Nvidia card that can serve the purpose as well as the RX 6800XT – as an example, the RTX 3060 combines an Ampere core with 12GB VRAM and would seem an excellent choice for a 3D content creator – if you can find one.

Games: First Person Shooters

We’ve grouped the results here in a per-game basis to simplify the comparisons, and to make it easy to focus in on the games or game types that interest you.

1. Call of Duty: Warzone

Firstly, looking at Call of Duty Warzone we can see that the two cards are relatively evenly matched: We run a full 5-6 minute Battle Royale, but against bots as this is the only way to make the benchmark a consistent and repeatable test. Settings are at the ‘High’ preset. The RX 6800XT has a very slight advantage at 1080p, but just 10 FPS at 250FPS isn’t a noticeable difference. At 1440p the two GPUs are evenly matched, and at 1440p Ultrawide they’re absolutely equal. At 4K the 3080 has a slight advantage, ahead 9 FPS at 128 FPS. We haven’t included other metrics to keep the chart legible, but the Minimum, Maximum, and 1%, and 0.1% lows are almost identical too. CoD Warzone doesn’t appear to have any particular preference and you’ll have an equivalent experience on either card. At 1080p you would be better looking at a lower-tier card and achieving nearly as high performance – an RTX 2060Ti achieves 220FPS at 1080p high settings, for example. 

2. Rainbow 6 Siege

Rainbow 6 Siege is a much faster-paced title and it shows in the benchmarks. Here, the RX 6800 XT takes a commanding lead at both 1080p and 1440p – it’s around 10% faster at 1080p and nearly 20% faster at 1440p. Initially, we felt this might be down to the rendering pipelines of the RTX 3080 being optimized for higher resolutions, but Hardware Unboxed and others have found interesting discrepancies in the Ampere cards performance when CPU limited: It appears their drivers have a higher overhead and as such it can dent the performance potential of the GPU. This game (and Flight Sim 2020) presents the most CPU-limited test we run, and as such, it may be one or both of these effects preventing the RTX 3080 from performing. It might also go some way to explaining that stellar 3DMark Fire Strike result at the start of this review. Again, at 1440p ultrawide resolution, we see that performance is about equal, and at 4K the RTX 3080 has taken the lead with 304Fps to 239 for the RX 6800 XT.

3. Doom Eternal

Finally, Doom Eternal is a well-optimized title using the Vulkan API and scales very well with both hardware and settings. We see a slight lead for the RX 6800 XT at 1080p at 409 FPS to 384, but at all other resolutions, the two GPUs perform identically on average. However looking at the underlying statistics, 1% and 0.1% lows are significantly lower for the RX 6800 XT indicating a less consistent experience and this occurs at every resolution. This isn’t something we see in the other First-person shooters where the results for these metrics are closely matched.

AAA Titles

Looking at the more demanding titles, this is where we expect to see both of these high end GPUs really shine.

Red Dead Redemption

Red Dead Redemption still ranks amongst the best looking and most demanding titles and again it uses the Vulkan API. Here we’ve run it at very high settings, near ultra, but with a couple of performance tweaks such as water being reduced to medium for a representative experience. Across the board these two GPUs perform near identically, exceeding 100 FPS at 1080p, close to 100FPS at 1440p, and exceeding 60FPS in 4K. Both provide a top tier experience, which is what you’d hope for flagship cards.

Shadow of the Tomb Raider

Shadow of the Tomb Raider always impresses us not only with its graphics but with its consistency when benchmarking – it reliably highlights very minor changes in system configuration or performance. At highest settings the numbers generated between these two cards are spookily similar, and not just the average frame rates, but the rest of the metrics as well. The experience on either card is identical, with one caveat: You can’t enable RTX Shadows or DLSS on the RX 6800 XT. However, in this title and with these GPUs this is a moot point as you don’t need either of these technologies for either visual improvements or performance. It’s a dead heat in this title.

Microsoft Flight Simulator 2020

Finally we’ll look at Microsoft Flight Simulator 2020.

This title continues to receive updates that fix bugs and improve performance, and it’s certainly a demanding test of a whole system. Our benchmark run is a four-minute AI piloted flight from LaGuardia over Manhattan at 2,500 – 3,500 ft – a worst-case scenario and highly demanding of the CPU as well as the GPU. 

It’s the CPU that limits the ultimate frame rate here, and even at low settings, 65FPS average is the absolute best performance that can be obtained. At more realistic and rewarding high settings, we can see how little the FPS scale with such powerful GPUs; at 1080p, 1440p and even 1440p ultrawide we achieve the same results, at our around 60FPS. We’re mostly CPU limited here. At 4K we start to see the impact of a demanding resolution AND settings even on these cards: High settings see 45-47FPS on both, whilst ultra settings cut performance to just over 30 FPS for both cards. The inconsistencies at lower resolutions are caused by CPU limitations, and not indicative of the relative performance of the cards, but we did have trouble extracting performance from the 6800XT at ultra settings and lower resolutions – it was reluctant to exceed 45FPS despite driver updates and tweaking. Complex games and hardware sometimes deliver these anomalies. At higher resolutions, it performed fine, and it’s likely a minor setting adjustment would have resolved the issue but invalidated the controls to our benchmark. 

The lesson here is how important balancing the whole system to the target resolution is in Flight sim 2020, as well as settings needing to be tweaked for optimum performance. Both of these cards are capable of playing MS Flight 2020 at high resolutions, a mix of high and ultra settings, and both are more than capable of an acceptable VR experience as well.

Related: Flight Simulator 2020 Performance Analysis: Get the most out of MS Flight 2020

Conclusions

As you’d hope and expect, both of these GPUs turn in an excellent performance. There’s nothing they won’t run well. Having spent over a month of tenting and general use with both of these cards, they are truly equal in performance in our opinion. There are some games and certain corner cases where one outperforms the other but on balance, the differences are infrequent and slight. 

It’s really in feature set and versatility where we see the RTX 3080 offer more to a prospective buyer. RTX and DLSS may or may not be the next big thing for gaming, but it’s nice to have the option and in many titles, it just isn’t there if you’ve got the AMD card. We hope that developers will use the DirectX 12 ultimate API to bring hardware Ray tracing to the broadest audience possible, rather than sticking to Nvidias’ ‘RTX’ walled garden, whilst AMD need to develop and deliver their ‘Super Resolution’ functionality to the FidelityFX suite, to challenge Nvidia in dynamic upscaling. 

Users who need a GPU for more than just gaming: for 3D content creation or data analysis, or to accelerate rendering – should look closely at benchmarks specific to their usage case. The RX 6800 XT performs better at some functions, and if you need a very high VRAM capacity then this or the 6800 is the cheapest way to get 16GB. It also offers better compatibility in Hackintosh and some Linux applications. However, for most tasks, the versatility of CUDA and Tensor cores combined with the higher bandwidth memory access make the 3080 a great choice. It will come down to what exactly you need to do.

Finally, the choice may well be out of your hands: If you see an opportunity to buy either of these GPU’s at a non-ridiculous price, and need a high-end gaming GPU, then either offers more than satisfactory performance at the highest resolutions and settings and will stand the test of time.

The post Nvidia RTX 3080 vs AMD RX 6800 XT: Benchmark Comparison (Real World Tests) appeared first on PremiumBuilds.