Linux 6.6 Delivers Some Impressive Gains For AMD EPYC 9754 "Bergamo" Server Performance

[skeevy420]

Unless you have the best golden sample out there, - 30 is probably not stabile. I have had errors with - 20 in prime95 and am now back to -15 which I haven't tested extensivey.

All core was stabile for me with - 20 but when i tested Single thread with and w/o HT i got errors. I recommend to test a few different threaded scenarios.

I've ran both prime95 and cinebench they've never crashed at -30, only higher, and my board has presets for up to -40. My CPU temp tops out at 85C under full benchmark load. MSFS2020 only gets it up to 65C.

7800X3D with a Noctua NH-U12S. Fucking beast of a CPU.

[Dr_ST]

Just to be sure, it is not something changed in other scheduling parts like 100Hz by default instead of 250Hz or 1000Hz that some distributions choose. I trained my students with this explanation in mind, a simple compilation change can bring a lot of performance improvement (server parameters versus desktop ones).
Are these parameters equal?

[unwind-protect]

Any chance to include FreeBSD in one of these scheduler-oriented benchmark runs?

[Michael]

Any chance to include FreeBSD in one of these scheduler-oriented benchmark runs?

Sometime maybe for a special article... FreeBSD doesn't run all of the tests that can be done on Linux. Not sure if I looked to see if the current FreeBSD release even works out on AMD Titanite reference server. And with being bottlenecked by one AMD SP5 server at the moment wouldn't tie it up another extra couple of day(s) for FreeBSD tests when it doesn't attract many readers. But sometime on some system can do a FreeBSD comparison.