AVX Optimization

dameyer's Avatar

dameyer

17 Jun, 2019 11:42 PM

Hi,

Anyone know what compiler was used to build geekbench 4? Testing an E5 v3 and Skylake processor I don't see the performance gains I would have expected if AVX optimization were included during the compile. May be some off settings in the BIOS but I expected to see the 6148 ripping the 2690 on all tests, especially multicore.

Thank you,
Doug

  1. Support Staff 1 Posted by John on 18 Jun, 2019 04:41 AM

    John's Avatar

    SGEMM and SFFT are the two workloads that are optimized to take advantage of AVX, AVX2, and AVX-512 instruction sets. The other workloads will only use SSE2 and SSE4 as generated by the compiler.

  2. 2 Posted by dameyer on 18 Jun, 2019 04:12 PM

    dameyer's Avatar

    I am in awe! Thank you for the fast and incredibly concise answer. One follow on question, For multicore tests, is the test using all available threads (Hyperthreads) or a reduced set? Some codes scale poorly as the cores in use increase. Just curious if you restrict the number of cores and/or restrict only to physical cores. I look forward to testing this on a Linux host.

    Thank you

Reply to this discussion

Internal reply

Formatting help / Preview (switch to plain text) No formatting (switch to Markdown)

Attaching KB article:

»

Attached Files

You can attach files up to 10MB

If you don't have an account yet, we need to confirm you're human and not a machine trying to post spam.

Keyboard shortcuts

Generic

? Show this help
ESC Blurs the current field

Comment Form

r Focus the comment reply box
^ + ↩ Submit the comment

You can use Command ⌘ instead of Control ^ on Mac