Intel's Clear Linux Helping AMD EPYC Genoa Hit New Performance Heights

[phoronix]

Phoronix: Intel's Clear Linux Helping AMD EPYC Genoa Hit New Performance Heights

Similar to the great results of Intel's performance-optimized Clear Linux on the Ryzen 9 7950X, making use of Clear Linux on the new 4th Gen EPYC "Genoa" processors also helps in maximizing performance for these AVX-512 server processors. Here are some initial benchmarks.

https://www.phoronix.com/review/epyc-genoa-clear

[Volta]

Ubuntu is terrible waste of CPU performance.

[mdedetrich]

I wonder why that 320 thread limit is there

[ms178]

And yet there are still some developers questioning the usefulness of system tuning and performance optimizations in general. They keep repeating the "O2 is enough for everybody" and "AVX2/AVX512 is a waste of silicon" mantra. I think this review proves it again that there are major gains to be had in certain workloads, and to be frank, I'd expect this performance work to be done by upstream developers and other mainstream distros, too, to provide a better out-of-the-box experience for their users.

Good job, Intel.

[ILoveGoulash]

As an alternative, I believe Gentoo with "-O3 -march=native", USE="lto pgo" and https://github.com/InBetweenNames/gentooLTO (if you want to go the last mile) should do.

[arjan_intel]

I wonder why that 320 thread limit is there

It was an oversight; we set that a couple of years ago and more or less forgot about it since we never hit that limit... the phoronix article is the first complaint we got basically.
I just bumped it way up to not have to think about it for another couple of years after this ;-)

[NeoMorpheus]

Intel engineer: Sorry for this bug, brb with a proper “fix”.

[pipe13]

Speaking of NWChem and oneAPI, have you considered benchmarking gfortran against /opt/intel/oneapi/compiler/latest/linux/bin/intel64/ifort ? I recently dragged it out to run some legacy fortran rife with computed goto's that gfortran (still) chokes on, and found ifort quite easy to configure and use.

[Michael]

Speaking of NWChem and oneAPI, have you considered benchmarking gfortran against /opt/intel/oneapi/compiler/latest/linux/bin/intel64/ifort ? I recently dragged it out to run some legacy fortran rife with computed goto's that gfortran (still) chokes on, and found ifort quite easy to configure and use.

Has ifort added compatibility against any of the common Clang and GCC flags? At least with DPC++ early on was going to benchmark that, but the various compiler flag differences and the time involved for ensuring consistency made it not worthwhile.

[JEBjames]

Michael

Typo page 1

"theiru pdates" should be "their updates"