Hacker GitHub

← All repos

geerlingguy/top500-benchmark

View on GitHub ↗ ★ 429 Jinja

Automated Top500 benchmark for clusters or single nodes.

8 mentions on Hacker News by 4 people: 0 stories, 8 comments · first 2023-01-24 · latest 2026-08-09

Mentions per year 8 mentions · busiest year 2024 (4)
20082026
AllStoriesComments NewestTopOldest
  1. …Also I was unable to put more than 8 processes in total. [2] The original m1 max benchmark: https://github.com/geerlingguy/top500-benchmark/issues/4
    comment by dimask · 2026-08-09 · in Can Intel finally beat ARM on performance per Watt?
  2. …(FP64, HPL) is also beating all the M4/M3/M2/M1 systems I've tested besides the M4 Mac mini: https://github.com/geerlingguy/top500-benchmark#results I think the bigger news is Intel's made a chip that isn't just pushing clocks to hit some…
    comment by geerlingguy · 2026-08-08 · in Can Intel finally beat ARM on performance per Watt?
  3. All his tests are published here: https://github.com/geerlingguy/top500-benchmark?tab=readme-o... He does test some N100 machines. TL;DR: The M4 blows everything away, but isn't a general purpose machine. The N100 is slightly better than…
    comment by rstuart4133 · 2024-11-13 · in M4 Mac mini's efficiency
  4. …still. You mention actually measuring it, that's exactly what I've been doing in my top500 project: https://github.com/geerlingguy/top500-benchmark#results The RK3588 gets up to 4.64 Gflops/W, only beaten by a couple expensive Ampere Altra…
    comment by geerlingguy · 2024-04-27 · in Turing RK1 is 2x faster, 1.8x pricier than Pi 5
  5. …not Raspberry Pis. It burns 18.5W with CPU cores maxed out, producing 224 Gflops of compute (see: https://github.com/geerlingguy/top500-benchmark/issues/27 ). Efficiency-wise, it's a little more power efficient than at least last-gen…
    comment by geerlingguy · 2024-04-26 · in Turing RK1 is 2x faster, 1.8x pricier than Pi 5
  6. HPL Linpack was tested on a Q64-22 CPU ~650 Gflops https://github.com/geerlingguy/top500-benchmark/issues/19 Memory bandwidth, it's 8 channels of DDR4 on the Ampere vs 7960x 4 channels of DDR5. AMD likely has GFLOPS in the low 1000s if I…
    comment by dylanetaft · 2024-03-05 · in Ask HN: ARM workstations for homelab CPU based AI Inference testing?
  7. …ways to calculate (and benchmark) flops, the way I'm doing it is with this open source project: https://github.com/geerlingguy/top500-benchmark — which can use OpenBLAS, Blis, or ATLAS, and I've tried all three un-tuned. I also worked with…
    comment by geerlingguy · 2023-10-27 · in Everything I've learned building the fastest Arm desktop
  8. Not super fast but efficiency is okay: https://github.com/geerlingguy/top500-benchmark#results
    comment by geerlingguy · 2023-01-24 · in Hyperscale in your Homelab: The Compute Blade arrives