'Clean' vs 'dirty' Windows 11 performance was benchmarked, results may surprise you

Back in February earlier this year, we benchmarked Windows 11 24H2 vs 23H2 performance on a clean-installed PC following one for an in-place upgrade to see if one or the other was the better way to upgrade.

According to Microsoft, a clean-installed Windows 11 PC can boost performance in a rather unconventional way thanks to its Smart App Control (SAC) feature.

Meanwhile, users often wonder if a reinstall or reset of Windows could help them gain performance. The idea is that the bloat and junk that accumulates over time may be holding back the hardware resources. As such, to see if a clean PC would be better compared to a 'dirty' one, tech news site PCWorld tested the performance across some games and other applications.

On the non-gaming tests, we have Photoshop first across a 9950X3D (16-core) and 9800X3D (8-core), and the results are interesting. The Clean 9950X3D system was faster, while the Dirty 9800X3D fared better. Overall, though, one could say the scores were within margins of error.

Up next, a couple of rendering benchmarks were presented, first, the synthetic Cinebench 2024, and then followed by the real-world Blender.

The clean systems did better on both Cinebench and Blender, which shows that perhaps when all cores and threads are at play, a reinstalled, fresh PC could be a better choice. However, keep in mind that even the biggest victories seen aren't something you will likely notice unless you are paying attention.

Up next, in games, the clean and dirty systems swapped positions in the case of Cyberpunk 2077, similar to what we saw in Photoshop above.

In F1 2024, the gaps were more pronounced with the 1% lows and even the average FPS showing higher differences.

Finally, in Fortnite, the clean PC exhibited significantly better 1% low framerates in the case of the 9950X3D.

Keep in mind, though, that Windows is known to be inconsistent when it comes to performance, and even run-to-run variances can be unexpectedly wider on occasions, which makes it difficult to conclude definitively in the case of such testing. Also, the 'dirty' PC that was measured here is a year old, so it is possible that we could be seeing bigger differentials had it been an even older system.

Source and images: PCWorld