7 ms·

Think you are mistaken. The M4 beats the Ryzen AI 365 in both single and multicore benchmarks

satellite2 1y ago
M4 23884 365 29861 https://www.cpubenchmark.net/cpu.php?cpu=Apple+M4+10+Core&id=6040 https://www.cpubenchmark.net/cpu.php?cpu=Apple+M4+10+Core&id... https://www.cpubenchmark.net/cpu.php?cpu=AMD+Ryzen+AI+9+365&id=6180 https://www.cpubenchmark.net/cpu.php?cpu=AMD+Ryzen+AI+9+365&...
- matdehaast 1y ago
  Interesting, on Geekbench they have very different scoring 365: 2515/12552 M4: 3763/14694 https://browser.geekbench.com/processors/amd-ryzen-ai-9-365 https://browser.geekbench.com/processors/amd-ryzen-ai-9-365 https://browser.geekbench.com/v6/cpu/11020192 https://browser.geekbench.com/v6/cpu/11020192
  - lostmsu 1y ago
    395 is still higher on Geekbench at 2781/17644
- FootballMuse 1y ago
  Passmark is an outdated benchmark not well optimized for ARM. Even so the single thread marks are 3864 (AI365) vs 4550 (M4) OTOH, Geekbench correlates (0.99) with SPEC standards, the industry standard in CPU benchmark and what enterprise companies such as AWS use to judge a CPU performance. https://medium.com/silicon-reimagined/performance-delivered-a-new-way-part-2-geekbench-versus-spec-4ddac45dcf03 https://medium.com/silicon-reimagined/performance-delivered-...
  - satellite2 1y ago
    I see you are citing a 6 month old post which itself isn't really well sourced isn't really reaching consensus and doesn't have a definitive answer. https://news.ycombinator.com/item?id=43287208 https://news.ycombinator.com/item?id=43287208 The article in question doesn't mention subpar ARM optimizations.
    - aurareturn 1y ago
      [flagged]
  - astrange 1y ago
    Hmm, why would you need to optimize a benchmark for something? Generally it's the other way round.
    - aleph_minus_one 1y ago
      > Hmm, why would you need to optimize a benchmark for something? Generally it's the other way round. It had always been both ways. This is why there exist(ed) quite a lot of people who have/had serious thoughts whether [some benchmark] actually measures the performance of the e.g. CPU or the quality of the compiler. The "truce" that was adopted concerning these very heated discussions was that a great CPU is of much less value if programmers are incapable of making use of its power. Examples that evidence the truth of this "truce" [pun intended]: - Sega Saturn (very hard to make use of its power) - PlayStation 3 (Cell processor) - Intel Itanium, which (besides some other problems) needed a super-smart compiler (which never existed) so that programs could make use of its potential - in the last years: claims that specific AMD GPUs are as fast as or even faster than NVidia GPUs (also: for the same cost) for GPGPU tasks. Possibly true, but CUDA makes it easier to make use of the power of the NVidia GPU.
      - astrange 1y ago
        > - Intel Itanium, which (besides some other problems) needed a super-smart compiler (which never existed) so that programs could make use of its potential Well, no such thing is possible. Memory access and branch prediction patterns are too dynamic for a compiler to be able to schedule basically anything ahead of time. A JIT with a lot of instrumentation could do somewhat better, but it'd be very expensive.

M4 23884 365 29861 https://www.cpubenchmark.net/cpu.php?cpu=Apple+M4+10+Core&id=6040 https://www.cpubenchmark.net/cpu.php?cpu=Apple+M4+10+Core&id... https://www.cpubenchmark.net/cpu.php?cpu=AMD+Ryzen+AI+9+365&id=6180 https://www.cpubenchmark.net/cpu.php?cpu=AMD+Ryzen+AI+9+365&...

matdehaast 1y ago
Interesting, on Geekbench they have very different scoring 365: 2515/12552 M4: 3763/14694 https://browser.geekbench.com/processors/amd-ryzen-ai-9-365 https://browser.geekbench.com/processors/amd-ryzen-ai-9-365 https://browser.geekbench.com/v6/cpu/11020192 https://browser.geekbench.com/v6/cpu/11020192
- lostmsu 1y ago
  395 is still higher on Geekbench at 2781/17644
FootballMuse 1y ago
Passmark is an outdated benchmark not well optimized for ARM. Even so the single thread marks are 3864 (AI365) vs 4550 (M4) OTOH, Geekbench correlates (0.99) with SPEC standards, the industry standard in CPU benchmark and what enterprise companies such as AWS use to judge a CPU performance. https://medium.com/silicon-reimagined/performance-delivered-a-new-way-part-2-geekbench-versus-spec-4ddac45dcf03 https://medium.com/silicon-reimagined/performance-delivered-...
- satellite2 1y ago
  I see you are citing a 6 month old post which itself isn't really well sourced isn't really reaching consensus and doesn't have a definitive answer. https://news.ycombinator.com/item?id=43287208 https://news.ycombinator.com/item?id=43287208 The article in question doesn't mention subpar ARM optimizations.
  - aurareturn 1y ago
    [flagged]
- astrange 1y ago
  Hmm, why would you need to optimize a benchmark for something? Generally it's the other way round.
  - aleph_minus_one 1y ago
    > Hmm, why would you need to optimize a benchmark for something? Generally it's the other way round. It had always been both ways. This is why there exist(ed) quite a lot of people who have/had serious thoughts whether [some benchmark] actually measures the performance of the e.g. CPU or the quality of the compiler. The "truce" that was adopted concerning these very heated discussions was that a great CPU is of much less value if programmers are incapable of making use of its power. Examples that evidence the truth of this "truce" [pun intended]: - Sega Saturn (very hard to make use of its power) - PlayStation 3 (Cell processor) - Intel Itanium, which (besides some other problems) needed a super-smart compiler (which never existed) so that programs could make use of its potential - in the last years: claims that specific AMD GPUs are as fast as or even faster than NVidia GPUs (also: for the same cost) for GPGPU tasks. Possibly true, but CUDA makes it easier to make use of the power of the NVidia GPU.
    - astrange 1y ago
      > - Intel Itanium, which (besides some other problems) needed a super-smart compiler (which never existed) so that programs could make use of its potential Well, no such thing is possible. Memory access and branch prediction patterns are too dynamic for a compiler to be able to schedule basically anything ahead of time. A JIT with a lot of instrumentation could do somewhat better, but it'd be very expensive.