Intel’s new generation of CPUs was released last month including the core i7-12700K. We’ve been given one to test and review. In this article, we’ll put it through its paces against the flagships from the last year to see how it measures up.
Intel has been lagging behind in the CPU wars for a couple of generations now. The 11th generation failed to challenge AMD’s Zen 3 line up and the 10 core i9-10900K is the last true powerhouse they released, now 18 months old.
To remedy this Intel have redefined CPU architecture, releasing the 12th generation, known as ‘Alder Lake’ with a hybrid design with both powerful P-cores for performance, and more efficient e-cores. This apes ‘big-little’ design CPUs found on mobile devices where efficiency is king, but we still want some high-performance cores for demanding tasks.
This CPU is fabricated at 10nm, which should improve efficiency and lower power use. Meanwhile, there are 8 P-cores on the 12700K, which have hyperthreading and can hit 4.9GHz all core speeds, and 4 e-cores too, which clocks at 3.9GHz maximum and lack hyperthreading. That makes this a 12 physical, 20 logical core CPU. Backing it up it’s got 25MB L3 cache, Intels UHD770 integrated graphics, and the K specification means this CPU is unlocked, so it can be tweaked for performance on Z690 chipset motherboards.
This CPU is vital for Intel to stamp their authority on the enthusiast CPU market, so we’re eager to find out what it can do.
Test methodology and System
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 CPUs under test. We’ve tested using an MSI Tomahawk Z690 Motherboard for the i7-12700K, A Z590 ROG Maximus XIII Hero for the 10th and 11th gen Intel CPUs, and the MSI Mortar B550 for the Ryzen 5800X.
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 set-up, according to the manufacturer’s intentions, but with the automatic optimisations in place. It’s also the default behaviour of the MSI Tomahawk Z690.
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.
We tested primarily with a Noctua NH-D15S cooler, but the performance was also verified with an Arctic Liquid Freezer II 240mm AIO. Thermal throttling was not encountered in any of the tests presented in this review.
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.
So, let’s dig into our results!
1. Synthetic Tests
Cinebench R20
Cinebench R20 allows us to test multicore or single-core 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. The i7-12700K Clearly brings its core advantage to this test, with 12 physical cores overwhelming the 10 cores of the 10850K. Running a single-core test demonstrates the performance of a single P-Core: The score of 737 points is a clear 100 points above that of the other three CPUs under test. Our main regret here is not having a 12 core 5900X available for test: No doubt it would be a close-run battle here for the multi-core crown.
Blender
Using Blender to render a couple of scenes, we get a sense of the rendering performance of these CPUs. This test is highly multithreaded, using all cores to maximum capacity until the workload is complete.
Note that shorter bars are better indicating less time taken: In this test, we can see that for the ‘Classroom’ render, the i7-12700K is a full 100 seconds faster to complete this workload than the next fastest CPU, the Ryzen 5800X. In the shorter BMW27 test, the Alder lake CPU is 30 seconds faster than the second-fastest CPU, the i9-10850K.
We feel obliged to point out that we’re using this as a synthetic test of the CPUs, and if you’re actually looking to accelerate 3D rendering an NVidia GPU will complete the task in a fraction of the time of even the 12th Gen Intel CPU here.
Clearly, the i7-12700K is very potent in multi-core workloads, with only the Ryzen 9 CPUs and the i9-12900K able to challenge it. It comfortably wins every test in this section.
3D Mark
Using 3D Mark we focus 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 i7-12700K stamps its authority on these tests as well, making significant gains over every other CPU on test. Just as in the other synthetic benchmarks, it’s the clear winner.
2. Game benchmarks
We ran our gaming benchmarks at 1080p and high settings 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.
Rainbow 6 Siege
Rainbow 6 Siege has an inbuilt benchmark which we’ve found very consistent.
In this benchmark, the i7-12700K turns the synthetic performance results into tangible performance gains, with 80FPS more than the 5800X, and more than 100FPS more than the flagship Intel 10th and 11th generation CPUs.
Doom Enternal
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:
This test initially showed the Ryzen 5800X beating the 12700K by a small amount: That’s an interesting result given the apparent single-core advantage of the Intel CPU. Brief analysis showed that Doom eternal is one of the games that Windows 11 struggles with on Alder Lake, so a switch back to Win 10 and a re-test showed the 12700K improving to the tune of 10fps average. At 380 FPS the performance is no slouch on either, but the 8 Core Zen 3 CPU still holds its own here. This result also highlights the challenges of a brand new platform and a new Operating System – performance refinements will continue as the operating system matures and better allocates tasks on this complex CPU.
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 swings back to the i7-12700K’s favour, with a clear 40FPS advantage over the other CPUs. Note we have isolated CPU performance here, so this isn’t indicative of actual FPS which will be GPU limited.
Red Dead Redemption 2
Red Dead Redemption 2 is another strong showing for the Ryzen 5800X.
Again it’s surprising to see the Ryzen 5800X doing well against the 12700K, with just a few FPS to the new CPUs favour. It’s possible we’re finding the limits of even an RTX 3080 at 1080p ultra settings, and whilst lower settings might show wider gaps we think it’s more interesting to demonstrate how close these CPUs can be ‘in the real world’. We re-ran this benchmark in Windows 10 and Windows 11 and found no appreciable performance difference, so this isn’t a case of the operating system limiting the new CPU architecture.
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. We’ve omitted the i9-11900K here as recent game updates have invalidated older testing with that CPU.
Here the i7-12700K is again the best performing CPU on test, using that spectacular single-core speed to deliver a 107FPS average. Note that core count doesn’t matter here, you can disable the 5800X or 10850K to 6 cores and obtain the same results. This test is all about cache size, and single-core speed and the 12700K has both in spades. We’ve got tonnes more in-depth testing on this game which will form a separate article, so if this sim is your focus you’ll want to keep an eye out for that. However, as a spoiler, the 12700K is absolutely the best option for this Simulator right now.
Gaming performance conclusions
Our game testing sees the i7-12700K either match or beat every comparable CPU in gaming. The Ryzen 5800X runs it pretty close in a couple of titles, however, in others we see a commanding 10% or so FPS lead. We’ve purposefully run these tests at more representative settings, do demonstrate rather than overstate the differences you’ll find between these CPUs.
Nonetheless, the result here is clear: At $400 The i7-12700K beats the Ryzen 5800X, and the outgoing Intel flagships. Given what we know of the 5900X and 5950X, where their performance in games is largely dependent on that same single-core speed as the 5800X, they don’t offer any compelling advantage in gaming except for in a few specific titles.
3. Memory Speed Scaling
RAM is the hot topic of Intels 12th Generation, since depending on your choice of motherboard you can use either DDR4 or DDR5 RAM. The newer specification remains very expensive and hard to find, whilst performance benefits outside of very specific tasks aren’t clear cut. We’ve tested with DDR4 Ram throughout this review: We feel it’s what the bulk of people will choose for this generation, particularly with the more sensibly priced i7-12700K.
However, the message persists that ‘intel doesn’t scale with RAM speed like Ryzen’ so we wanted to find out if the i7-12700K was sensitive to RAM speeds.
To illustrate this, we ran the Shadow of the Tomb Raider Benchmark in a variety of speed configurations:
These tests cover the spectrum from ‘getting it wrong’ with default JDEC specification RAM, such as you’d encounter if you failed to set XMP, through to commonly available kits from 3200Mhz and 3600Mhz CL16, up to overclocked and somewhat optimised DDR4 RAM at 4000MHz Cl15-16-16-32 in Gear1.
You can see there is relatively consistent performance scaling as RAM latency decreases, but it’s not dramatic. We use Shadow of the Tomb Raider for this demonstration because it is responsive to RAM tweaking, many situations are not. Nonetheless, we can see that with a relatively affordable 3600MHz CL16 RAM kit, we have the bulk of performance on offer with minimal investment in both money and time. It remains our pick for the best RAM option for high-performance Intel CPUs into the 12th Generation. That said, we found memory overclocking easy and fun on this platform: If you do want to tweak, we can recommend a high-performance B-Die kit, and no doubt timings could be significantly optimised from those used to demonstrate this result.
We have separate content coming expanding on this aspect of Alder lake CPU performance.
4. Power and Thermals
Power draw and the consequent heat output has long since been the cost of high performance on Intel’s CPUs. We ran tests to explore this on the i7-12700K. We opted for the popular NH-D15S Cooler to examine the performance of a top tier air cooling solution on this CPU.
This CPU Maintains the Intel standard of a 190W PL1 for the duration of this test. Core speeds remain at 4.7GHz throughout – and did not throttle even in an extended 10-minute test. CPU temperature is maintained at a thoroughly manageable 79 °C. We repeated this test with an Arctic Liquid Freezer 240mm AIO and obtained the same results – both coolers were plenty capable of handling this CPU at default settings.
We ventured into overclocking, adding 1000 points to our Cinebench R23 Score with a 5GHz P-core and 4GHz e-core target. Results came at the expense of a 240W Power draw, and temperatures in the mid 90’s despite a -50mV undervolt. If you do intend on overclocking this CPU, we’d advise a 280mm or 360mm AIO as a minimum. That said it was thoroughly manageable and entertaining to see an Intel CPU respond to overclocking positively once again.
Who is this CPU for?
The i7-12700K suits a broad range of workloads and needs. It’s the sweet spot for high-end gaming, content creation and computational workloads. Whilst the Ryzen 9 CPUs offer more physical cores, the times when they are brought to bear on most peoples tasks are minimal. Meanwhile, the faster individual core speeds of the 12th generation assist much more of the time, delivering higher FPS in gaming, snappier processing in adobe apps and other tasks of that nature. The iGPU is also a bonus to many workloads, accelerating transcodes and transforms for video editors and digital artists.
The i5-12600K is a very valid option at around $100 less, for those workloads if you’re on a budget or for gamers who don’t need 8 P-Cores. The i9-12900K adds 4 more e-cores and remains the preserve of the high-end enthusiast. Most people will be better off saving money with an i7-12700K and buying a better GPU, more SSD space or more RAM.
The imminent release of the non-K CPUs also looks compelling. The first test of the i7-12700 show it performing incredibly close to the K variant: It may well be a sensible choice to keep budgets in control. Meanwhile, the i5-12400 looks set to become the new budget gaming champion, eclipsing the performance of the Ryzen 5600X in a $200 product.
AMD is now left somewhat out in the cold: Whilst the platform costs of the Zen 3 CPUs are lower, the 5800X at $400 still makes little sense against a $400 i7-12700K, and at $300 the i5-12600K matches or outperforms it an offsets the higher motherboard cost. AMD have a response in the pipeline in early 2022 with the ‘stacked V-Cache’ version of the 5800X, the 5800X3D CPUs, so it will be interesting to see how much 92MB total cache can make up the performance gap. The Ryzen 9 CPUs are still significantly more expensive, and their core counts don’t help most users nearly as much as the faster cores of Intels 12th Gen. You need a very specific workload for a Ryzen 9 to be the best choice of CPU right now.
However, if you’re sitting there with an Intel 10th generation or a Ryzen Zen 3 CPU – I wouldn’t take the hype around this release as a cue to upgrade. This CPU is a good step forwards, but it’s not enough of a leap to warrant a platform change from those relatively recent and still high-performance CPUs unless you’re suffering poor performance due to CPU limitations.
Conclusion
In conclusion, it has been nice to be impressed by an Intel CPU. The i7-12700K is an absolutely storming CPU and excels across a range of workloads, from heavily multithreaded productivity tasks to gaming. This i7 CPU happily beats the last 2 flagship Intel CPU’s, and it’s only challenged in multithreaded superiority by the Ryzen 9 CPUs and the current flagship i9-12900K.
This generation has righted many of the wrongs from the 11th generation: Power draw and temperature are once again sensible. Performance is outstanding. Where the i9-11900K felt like you had to work to extract performance from it, the i7-12700K willingly demonstrates its prowess.
This CPU does many things right, and for most people looking to build a PC now, this or the i5-12600K are the right choices. However, if these CPUs and the accompanying $200+ Z690 motherboards push you over budget, keep an eye out: Early 2022 will see the value options become available, the i5-12400 and i3 parts based on this platform, as well as more affordable B660 motherboards. On the evidence of these flagship CPUs, and given the dearth of budget AMD CPU options at the moment, we should see Intel regain a dominant position In the CPU market.