Hacker News new | ask | show | jobs
by squeaky-clean 1120 days ago
Main reason I know of is having a new CPU with separate P/E cores. The Windows 10 scheduler assumes all cores are identical in performance.
2 comments

Has anyone been able to come up with a real-world benchmark showing a performance difference? My understanding is that the effect of the new scheduler is almost unnoticeable.

In any event, no new OS features are worth dealing with nonconsensual ads popping up whenever I move the mouse the wrong way.

So the new P CPUs like i7-1260p etc. are not going to be effective on Win 10?
It will run but you won't get the full benefits of the split between performance and efficiency cores.

Looks like I was a bit wrong in my earlier comment, they have recently updated Windows 10 to support the concept of P/E core differences, but the way they assign processes to specific cores in Win10 is very unsophisticated compared to Win11.

Windows 11 task scheduler talks directly to the Thread Director, a microcontroller added added on gen 12+ CPUs, while Windows 10 scheduler was modified after release to account for them through collaboration with Intel.

According to both MS and Intel, thread scheduling on 10 is not as optimized as on Win 11.

According to benchmarks (typically games), performance is roughly the same. Most wins go to Windows 11, some to Windows 10, but within negligible ranges of difference to where it doesn't matter on either. That was reason enough to stick with Windows 10 on my 13600k i5 build still.