News for hackers. Stuff that matters.
Re:Ask HN: Running production server on M1 mini? by gjsman-1000 on Monday December 06, 2021 @09:14AM (#29459935)
I’d suggest investigating single core performance. If you have the money, buy an i9-12900K (slightly faster single-core than M1 but much hotter) and do some testing on that. If my theory is correct, performance will be even better.
Re:Ask HN: Running production server on M1 mini? by groundthrower on Monday December 06, 2021 @09:21AM (#29460010)
We have examined that as well, last week we tried a AMD 5950X which has half the amount of cores but much better single core performance - the result was still at 60% of the Epyc performance
Re:Ask HN: Running production server on M1 mini? by gjsman-1000 on Monday December 06, 2021 @09:23AM (#29460033)
What was the M1 % relative to your Epyc?
Re:Ask HN: Running production server on M1 mini? by groundthrower on Monday December 06, 2021 @09:28AM (#29460094)
Roughly 10% faster
Re:Ask HN: Running production server on M1 mini? by CobaltFire on Monday December 06, 2021 @01:53PM (#29463984)
Have you investigated memory constraints?

Ryzen is 2 channels; Epyc is 4-8 (depending on CPU). M1 has that stupidly fast/wide setup.

If your Epyc is one of the 4 channel optimized SKUs or is only running in 4 channel mode, you would get pretty close to the quoted ratios on a memory bandwidth test.

Correlation, not causation, but worth looking into.

Re:Ask HN: Running production server on M1 mini? by fomine3 on Monday December 06, 2021 @09:40PM (#29468590)
Also check Node per Socket (NPS) settings on EPYC
Re:Ask HN: Running production server on M1 mini? by gjsman-1000 on Monday December 06, 2021 @09:38AM (#29460207)
HN makes us wait for replies… so if we need to continue this further I’m open at [email protected] .

My next question would be if you ran the 12900K in dual-channel memory.