Missing metric set for GH200

Developer Tools Nsight Systems Profiling Linux Targets
1

Hi,

I do not see the metrics-set for GH200 on a recent NSys version. Could you please let me know if there is a manually specified file that I can use for the metrics set on the GH200 counters while this Nsys aspect is in development?

(base)$  nsys profile --gpu-metrics-set=help
Possible --gpu-metrics-set values are:
        tu10x        : General Metrics for NVIDIA TU10x (any frequency)
        tu11x        : General Metrics for NVIDIA TU11x (any frequency)
        ga100        : General Metrics for NVIDIA GA100 (any frequency)
        ga10x        : General Metrics for NVIDIA GA10x (any frequency)
        gh100        : General Metrics for NVIDIA GH100 (any frequency)
        ad10x        : General Metrics for NVIDIA AD10x (any frequency)
        ga10b-gfxt   : Graphics Throughput Metrics for NVIDIA GA10B (frequency >= 10kHz)
        tu10x-gfxt   : Graphics Throughput Metrics for NVIDIA TU10x (frequency >= 10kHz)
        ga10x-gfxt   : Graphics Throughput Metrics for NVIDIA GA10x (frequency >= 10kHz)
        ad10x-gfxt   : Graphics Throughput Metrics for NVIDIA AD10x (frequency >= 10kHz)
        ga10x-gfxact : Graphics Async Compute Triage Metrics for NVIDIA GA10x (frequency >= 10kHz)
(base) $ nsys --version
NVIDIA Nsight Systems version 2024.1.1.59-241133802077v0
2

First of all, that version is from the beginning of this year, that isn’t recent in terms of chip support.

Secondly, may I inquire where you are using a GH200 chip?

3

@hwilper Could you please let me know which version of Nsight Systems includes gh200 support?

4

GH200 did not ship until Siggraph. The Nsys version shipped at Siggraph is 2024.5

5

Hi @hwilper , I have the update nsys version now but the metrics-set still does not include gh200. What am I missing here? Why is this the case?

$ nsys --version
NVIDIA Nsight Systems version 2024.5.1.113-245134619542v0

$ nsys profile --gpu-metrics-set=help
Possible --gpu-metrics-set values are:
        tu10x        : General Metrics for NVIDIA TU10x (any frequency)
        tu11x        : General Metrics for NVIDIA TU11x (any frequency)
        ga100        : General Metrics for NVIDIA GA100 (any frequency)
        ga10x        : General Metrics for NVIDIA GA10x (any frequency)
        gh100        : General Metrics for NVIDIA GH100 (any frequency)
        ad10x        : General Metrics for NVIDIA AD10x (any frequency)
        ga10b-gfxt   : Graphics Throughput Metrics for NVIDIA GA10B (frequency >= 10kHz)
        tu10x-gfxt   : Graphics Throughput Metrics for NVIDIA TU10x (frequency >= 10kHz)
        ga10x-gfxt   : Graphics Throughput Metrics for NVIDIA GA10x (frequency >= 10kHz)
        ad10x-gfxt   : Graphics Throughput Metrics for NVIDIA AD10x (frequency >= 10kHz)
        ga10x-gfxact : Graphics Async Compute Triage Metrics for NVIDIA GA10x (frequency >= 10kHz)
        file:<file name> : use metric set from a given file
6

@pkovalenko

7

The GH200 is a product that uses the gh100 chip architecture. Please use one of the GPU metric sets for gh100.

8

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.