Announcement

Collapse
No announcement yet.

AVX-512 Performance Comparison: AMD Genoa vs. Intel Sapphire Rapids & Ice Lake

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • AVX-512 Performance Comparison: AMD Genoa vs. Intel Sapphire Rapids & Ice Lake

    Phoronix: AVX-512 Performance Comparison: AMD Genoa vs. Intel Sapphire Rapids & Ice Lake

    With last week's launch of Intel's 4th Gen Xeon Scalable Sapphire Rapids server processors, Intel heavily talked up the shiny new accelerators and the big performance potential of AMX, but not really showcased and only heard through the grapevine was the improved AVX-512 implementation found with these new processors. With Sapphire Rapids there is reduced penalties from engaging AVX-512 -- and for some AVX-512 instructions, no longer any measurable impact -- compared to prior generation Xeon processors. In this article is a look at the performance for a wide variety of workloads with AVX-512 on/off not just for Sapphire Rapids but also for prior generation Ice Lake as well as AMD's new EPYC 4th Gen "Genoa" processors where they have introduced AVX-512 for the first time.

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    Just out of curiosity, how is listed AVX2 among the CPU's instructions?

    Comment


    • #3
      Originally posted by MorrisS. View Post
      Just out of curiosity, how is listed AVX2 among the CPU's instructions?
      AVX2 is still exposed. See screenshot for all the flags.
      Michael Larabel
      https://www.michaellarabel.com/

      Comment


      • #4
        Originally posted by Michael View Post

        AVX2 is still exposed. See screenshot for all the flags.
        many thanks, I've detected it.

        Comment


        • #5
          To quote the wisdom of the Simpsons: "Stop, stop, he's already dead!"

          Comment


          • #6
            It would be nice if -mprefer-vector-width=256 + avx512 was also benchmarked. For zens most of benefit comes actually from ISA extensions rather than greater vector width.

            Comment


            • #7
              coder As I remember our discussions from a couple of weeks back about AVX-512, power consumption and frequency behavior, we have more data now that confirm that old wisdoms no longer apply. For the newest generation, AVX-512 is a solid win in applications that can make good use of the new instructions or vector width.

              Comment


              • #8
                What a pity that Intel couldn't make AVX-512 work on Raptor Lake. Their competitive stance against Zen 4 would be much better if their desktop SKU could also profit that much from its use. I wonder if Intel tries to compete with their HEDT lineup against the higher core count parts of Zen 4. While it came down to pricing, that would be more exciting than bringing a 6-P-core notebook chip to the desktop.

                Comment


                • #9
                  Originally posted by ms178 View Post
                  What a pity that Intel couldn't make AVX-512 work on Raptor Lake. Their competitive stance against Zen 4 would be much better if their desktop SKU could also profit that much from its use. I wonder if Intel tries to compete with their HEDT lineup against the higher core count parts of Zen 4. While it came down to pricing, that would be more exciting than bringing a 6-P-core notebook chip to the desktop.
                  A single Sapphire Rapids tile would be dual channel... I know its not gonna fit on the consumer socket, but it sure would be interesting if it could.

                  Comment


                  • #10
                    Originally posted by Article
                    I'm left rather surprised that Intel hadn't more notably promoted their AVX-512 improvements with 4th Gen Xeon Scalable at launch
                    I'm not. Intel will only highlight the situations where they win by a country mile... and that basically means their new accelerator extensions which AMD cannot compete with, because AMD does not have them.

                    Comment

                    Working...
                    X