Windows 11 is official out, and users around the world are getting a taste of Microsoft’s next-gen OS. One of the primary differences between Windows 10 and 11 is the ability to distinguish between performance and efficiency cores (in Intel’s upcoming Alder Lake CPUs), and may as well be the sole reason it exists. As per tests, performance in Lakefield (1+4) generally improves across the board with Windows 11, but what about homogenous CPUs?
Luckily for us, German outlet ComputerBase quickly benchmarked Windows 11 on both Intel and AMD systems to gauge the performance improvements/decrements.
| Test | Windows 10 | Windows 11 |
|---|---|---|
| Application performance | ||
| Cinebench R20 single / multi | 599 / 4,361 points | 601 / 4,349 points |
| Cinebench R23 single / multi | 1,560 / 11,296 points | 1,541 / 11,266 points |
| Agisoft PhotoScan | 234.47 seconds | 235.88 seconds |
| Handbrake 1.4.1 | 89 seconds | 89 seconds |
| 3DMark Time Spy (Total / GPU / CPU) | 12.541 / 13.292 / 9.501 | 12.663 / 13.450 / 9.513 |
| 3DMark CPU Profile (Max / 1 Thread) | 6.341 / 1.001 | 6.406 / 996 |
| PCMark 10 total (3 sub-areas) | 7,717 (10,894 / 9,804 / 11,675) | 7,844 (10,654 / 10,408 / 11,812) |
| Bapco CrossMark total (3 sub-areas) | 1,686 (1,680 / 1,764 / 1,486) | 1,762 (1,768 / 1,776 / 1,707) |
| Performance in games | ||
| Test in Full HD, maximum details | Windows 10 | Windows 11 |
| F1 2020 FPS / 0.2 | 245.6 / 173.5 FPS | 245.0 / 172.6 FPS |
| Borderlands FPS / 0.2 | 128.6 / 96.4 FPS | 128.0 / 99.4 FPS |
| Gears Tactics FPS / 0.2 | 209.7 / 163.2 FPS | 198.8 / 152.0 FPS |
| Kingdom Come FPS / 0.2 | 96.0 / 42.2 FPS | 96.9 / 38.6 FPS |
| Metro Exodus FPS / 0.2 | 98.4 / 62.1 FPS | 93.6 / 61.0 FPS |
Unfortunately, there’s a small but consistent regression in performance upon switching to Windows 11, especially in gaming workloads. Although the deltas are close to negligible, I’d suggest sticking to Windows 10 for the time being, at least till Microsoft irons out all the issues with the new OS.
| Test | Windows 10 | Windows 11 |
|---|---|---|
| Application performance | ||
| Cinebench R20 single / multi | 651 / 10.189 points | 647 / 10.137 points |
| Cinebench R23 single / multi | 1,657 / 26,196 points | 1,663 / 26,168 points |
| Agisoft PhotoScan | 133.45 seconds | 135.01 seconds |
| Handbrake 1.4.1 | 45 seconds | 46 seconds |
| 3DMark Time Spy (Total / GPU / CPU) | 18.353 / 19.886 / 12.775 | 18.355 / 19.936 / 12.664 |
| 3DMark CPU Profile (Max / 1 Thread) | 12.438 / 995 | 12.102 / 993 |
| PCMark 10 total (3 sub-areas) | 9,009 (11,279 / 10,596 / 16,600) | 8,808 (10,899 / 10,332 / 16,468) |
| Bapco CrossMark total (3 sub-areas) | 1,627 (1,612 / 1,673 / 1,542) | 1,595 (1,551 / 1,715 / 1,392) |
| Performance in games | ||
| Test in Full HD, maximum details | Windows 10 | Windows 11 |
| F1 2020 FPS / 0.2 | 309.3 / 199.6 | 309.3 / 200.9 |
| Borderlands FPS / 0.2 | 170.3 / 127.5 | 169.7 / 131.7 |
| Gears Tactics FPS / 0.2 | 181.9 / 146.7 | 170.4 / 137.3 |
| Kingdom Come FPS / 0.2 | 83.6 / 36.3 | 80.8 / 36.0 |
Interestingly, popular benchmarking applications like 3DMark, PCMark, and Bapco (generally) see a notable performance uplift with Windows 11…wait for it…on Intel systems. On AMD, we see roughly the same amount of performance deficit with these benchmarking tools. The reason behind this discrepancy is unclear, but it should be fixed in the coming weeks. We’ll update you as the story develops and will be sure to let you know when to make the switch.
