flat assembler
Message board for the users of flat assembler.
Index
> Projects and Ideas > NUMA CPU and RAM benchmark. |
Author |
|
manusov 13 Aug 2023, 13:05
Processor and memory benchmark for Windows.
Pure assembler and open source. Cache and DRAM bandwidth and latency measurement. 32/64-bit supported. AVX512 supported. Source and executables: https://github.com/manusov/NCRBv2
_________________ Information belongs to Universe. |
||||||||||||||||||||||||||||
13 Aug 2023, 13:05 |
|
manusov 15 Aug 2023, 05:20
Scale (and optimal drawings location) for Y-axis can be selected (separately for Latency and MBPS, L1-L4 caches, DRAM). Some example in the ncrb.inf file.
_________________ Information belongs to Universe. |
|||
15 Aug 2023, 05:20 |
|
manusov 15 Aug 2023, 05:31
Additionally, Silent Mode option can be used for more stable results and clear lines,
but without realtime drawings.
_________________ Information belongs to Universe. |
||||||||||
15 Aug 2023, 05:31 |
|
CandyMan 15 Aug 2023, 10:24
Fun and interesting programme. Thanks.
|
|||
15 Aug 2023, 10:24 |
|
Furs 15 Aug 2023, 12:40
Very cool, works on Linux too with Wine.
|
|||
15 Aug 2023, 12:40 |
|
manusov 22 Oct 2023, 07:50
It's very interest make same test for GPU-initiated memory access
and measuring low-level performance for PCIe and Video Memory (GDDR). But too many device-specific factors for low-level GPU programming. Example below yet verified at NVIDIA RTX 3060 Ti only. https://github.com/manusov/LearningVisualStudio/tree/master/simple_test_compute_shader/GPU_BANDWIDTH https://github.com/manusov/LearningVisualStudio/tree/master/simple_test_compute_shader/GPU_LATENCY
_________________ Information belongs to Universe. |
|||||||||||||||||||
22 Oct 2023, 07:50 |
|
< Last Thread | Next Thread > |
Forum Rules:
|
Copyright © 1999-2025, Tomasz Grysztar. Also on GitHub, YouTube.
Website powered by rwasa.