Intel Core i9-12900K VS AMD Ryzen 9 5950X: Win 10 vs Win 11 | Benchmark | Test

Intel Core i9-12900K VS AMD Ryzen 9 5950X: Win 10 vs Win 11/ Benchmark/Test – Comparison of performance of Intel Core i9-12900K and AMD Ryzen 9 5950X in Windows 10 and Windows 11

Core i9 12900K vs Ryzen 9 5950X – Test in 10 Games

Today we will try to understand one delicate issue, the debate about which has not subsided lately: in which version of Windows are tests-not tests stronger, and in which direction should enthusiasts fight?

The main core of the debate is that the 12th generation of Intel processors, according to manufacturers, should work better in the new version 11 of the operating system, while Ryzen processors, on the contrary, experienced certain problems with it. And to compare results in different OS is somehow wrong.

The situation with the choice of a single version of the OS for comparisons is aggravated by constant updates, any tests can become irrelevant in the shortest possible time.

So, we update everything that is possible (windows, games, drivers) and proceed.

Intel Core i9-12900K VS AMD Ryzen 9 5950X: Test stand

As is already clear from the introduction, the main subject of study is the 10 and 11 versions of Windows. We will compare the behavior of the flagship Core i9-12900K and Ryzen 9 5950X processors. The more cores, the more opportunities to see the difference. Motherboards and all other components correspond to the level.

Both processors are in fixed overclocking. 5950X is still the same: 4.85 / 4.825 GHz on CCX.

12900K new – we continue to collect statistics on overclocking potential.

This copy is about the same in terms of the memory controller – 4100 MHz in Gear1. The ring tire is slightly worse. Its fixed maximum is 4400 MHz, but the cores themselves are a little more successful: at the same 5.2 / 4.1 GHz, this 12900K is able to show itself in LinX.

Tests in benchmarks

We will immediately tell you about some of the nuances that are very noticeable. In Windows 10, newer Intel APUs sometimes exhibit oddities. For example, testmem5 will only run on efficient E-cores. Similar maneuvers can be performed by archivers, rendering in a video editor. On Windows 11, the Tread director does not allow himself to do this and gives no reason to doubt his professional suitability.

Ryzen also has its own characteristics, but, on the contrary, in version 11 of the operating system.

The recommended chipset drivers in some cases allow you to transfer a low-threaded task to the correct core with the best boost, but at the same time, L3 cache in AIda64 synthetics does not behave very stable.

Without the L3 chipset driver installed, the cache seems to work better in the same non-ideal test, which is probably reflected in the results of sensitive gaming tests. A little, but still.

In Windows 10 on 12900 in a fixed overclock, the single-threaded Cinebench test still runs on the first core and with a reduced effective frequency, hence the result is worse than with the bus to the same 5.2 GHz. In the 11th such behavior is not observed.

The 5950x does not experience such problems and it does not matter in which version of Windows the squares are rendered. In a multi-threaded test, the difference is difficult to detect.

Read This Now:   Intel Core i7-9700KF Processor: Overview (Special)| Overclocking | Benchmark | Specs

Another benchmark – CPU-Z, despite the vague practical application of the scores, does not show such problems with a single-threaded load in 12900K; the task spins on the cores marked for the best boost, but a little more successfully in the 11th version of Windows than in the 10th.

For the 5950s, such a feature was not delivered. Only the first core is still loaded. However, for fixed overclocking, this is not a problem. It turns out that the difference between versions of Windows is small and only in a single thread.

In the Blender benchmark, as in Cinebench multithreading, there is essentially no difference. If you need to load all the available capacities, then the tricky distribution of flows will not noticeably affect the final result.

Let’s take a look at the CPU Profile test results from the 3DMark package. The multi-threaded result shows that the 12900K is indifferent to the OS version, while the Windows 10 version is preferable for Ryzen 9.

The single-threaded test does not bring any surprises. For 12900K, version 11 is a little more preferable, and 5950X as a whole makes no difference.

The other number of threads in the test does not show anything interesting, except for eight. Here, suddenly, there is an almost ten percent difference noticeable for the 5950x.

A closer look reveals a strange load distribution. Instead of using only physical cores, logical threads are loaded for some reason.

Moreover, installing the chipset driver corrects the situation a little, but not dramatically.

So, before gaming tests, there is an assumption about the advantage of Windows 11 in low-thread for 12900K and less dependence on single-thread OS for 5950X. But for Ryzen, in general, we should expect slightly less gaming performance on the new operating system, and in the case of tasks not for all cores, even strange situations are possible. At the same time, the full load of all available capacities should reduce the differences in Windows versions to a minimum.

Tests in games

Cyberpunk 2077, 1080p, ray tracing ultra preset, FOV 85, DLSS ultra performance, one ray tracing item off – all to reduce the load on the graphics card.

The overall load across all streams is quite high. There is no way to identify any patterns in the distribution of threads, but a slightly higher performance of 12900K in Windows 10 can be noted. For Ryzen 9, there is almost no difference in FPS, but there is a certain tendency for Windows 11 to use SMT more actively in terms of threads.

Dry numbers indicate a slight, but the advantage of the old version of the OS.

Far Cry 6, 1080p, ultra preset, DXR both settings enabled, resolution scale set to 0.5.

Everything is much more interesting here. In Windows 10 12900K, it experiences some oddities in load distribution, 6 P-cores are loaded, two are idle, but the efficient ones are all busy with some kind of work.

In Windows 11, the distribution looks more correct: each of the productive cores received one loaded thread, and each cluster of E-cores received a small load distributed over two cores (at least from the point of view of cache usage, this looks much more optimal).

Read This Now:   Afox GeForce GTX 750 Ti (4 GB) Review: Benchmark | Testing | GTX 1050 VS 750 Ti

The difference in FPS against the background of such differences in the distribution of threads is quite insignificant. Apparently, tasks on different threads are performed equally successfully both there and there, and their importance is not great.

For the 5950X, something interesting could not be seen. The FPS is also almost the same, but in contrast to 12900K, we have the expected barely noticeable drawdown in the 11th version of Windows.

If you look only at the final FPS numbers, then you won’t guess what is happening with the distribution of threads.

Starcraft2, 1080p.

The load on the cores of the second CSH for the 5950 was explained before the gaming tests by the lack of a chipset driver, but this does not have a noticeable effect on FPS due to the fixed overclocking and high frequency of all cores.

With 12900K everything is much more interesting again. In Windows 10, one and a half threads are familiar and performance has already been measured, but in version 11 of the OS, monitoring tools cannot catch the main thread. According to HWiNFO, it is always possible to register loaded 4 threads and their employment is far from the maximum. But this does not prevent to get the final best result.

Moreover, the difference between the versions of the operating system is the largest of those registered in comparison. Here are the cases: the oldest game and the biggest benefit of switching to a new Windows.

Shadow of the Tomb Raider, 1080p, maximum preset, to reduce the load on the video card, we had to not only overclock it, but also reduce the resolution modifier to 20%.

12900K does not let you get bored. In Window 11, E-cores are not used at all, hyper-threading is involved from the very start of the scene. In Windows 10, the start loads strictly on the thread on the core. The performance of these two approaches is comparable, the new operating system is almost imperceptibly ahead.

For the 5950, the situation is similar, but, on the contrary, a drop in performance is recorded in version 11 of the OS. And basically it is at the beginning of the third scene, in difficult places there is practically no difference.

If we evaluate only the FPS numbers, then the essence of the differences in behavior in different operating systems for the 12th generation of Intel processors is lost.

Total War Saga: Troy, maximum preset, unit size and grass quality are extreme, resolution modifier is 50%, anti-aliasing is disabled.

The higher the processor load, the smaller the differences between versions of Windows. Almost like in renderings – everything that is available is used, and there is no any freedom in the management of computing power. It is curious that only in the most difficult places the 12900K is slightly inferior to the 5950x, but on average it significantly outperforms its 16-core competitor in this game.

According to the final result, you can’t say that somewhere in general you can find a noticeable difference between operating systems. In Troy it doesn’t exist.

Read This Now:   AMD Ryzen 7 3800X Processor: Overview (Nice!)| Overclocking | Benchmark | Specs

Watch Dogs: Legion, 1080p, FOV 90, Ultra Preset, DXR Ultra, Resolution Scale 25%.

For the 12900K, we can see a more explicit move away from hyper-threading in Windows 11, which is contrary to the current trend. For 5950 this is less pronounced, but the overall impression is the same.

The efficient cores of the new Intel processor are used in both operating systems, and in general, the difference for this game can be traced only in the final result. It is curious that with all the current updates, Windows 10 is slightly better than Windows 11, even in rare stutters on Alder-lake.

Warzone, 1080p, esports settings.

In this online shooter, the situation is similar to the Legion, only the opposite is true. More clearly seen is the desire of the operating system to divide the task so that each core has one thread in Windows 10 and not with Intel, but, on the contrary, with Ryzen.

Maybe, of course, monitoring does not manage to catch all the nuances, but for now, this behavior can be noted. In Warzone 10, the operating system version is clearly preferred by both AMD and Intel.

Conclusion

What conclusions can be drawn from the example of these demonstration games? Windows 10 is indeed preferable for AMD processors, but it is unequivocal to say that the 12th generation of Intel feels more confident everywhere on a fresh operating system. There are cases, like with Starcraft, where Windows 11 is really useful, but the reverse is true with Warzone. Based on the combination of gaming and synthetic comparisons, we can conclude that the thread director behaves more adequately in Windows 11, and this is a really strong argument in favor of this operating system.

It is important to understand that comparisons are valid only up to significant updates, and they happen very often. For example, in our previous attempt to compare Windows 10 and 11, the more experienced system outperformed both Alder-lake and Zen processors. It is quite possible that in a month or even a week later the situation will be diametrically opposite.

New versions of applications may also produce changes in interaction with Intel APUs.

In our comparisons, we would like to keep the conditions level and not to push the processors running different operating systems. Testing in both versions causes an avalanche-like increase in labor costs and is also unacceptable. I would like to hear your opinion, which version is better to compare? At 10 or 11?

That’s all, until we meet again, light updates and good health!


Notice: ob_end_flush(): failed to send buffer of zlib output compression (1) in /home/gamefeve/bitcoinminershashrate.com/wp-includes/functions.php on line 5373

Notice: ob_end_flush(): failed to send buffer of zlib output compression (1) in /home/gamefeve/bitcoinminershashrate.com/wp-includes/functions.php on line 5373