Hi, I am running a ML workload and I would like to know the matrix sizes - m, n, k and the data types of these matrices. With nsys I am able to see the actual gemms but not the matrix size, transpose flags, and the data types of input/output matrices. How or which profiler would give out these info…

ML workload - getting matrix sizes from profiler

Accelerated Computing CUDA CUDA Programming and Performance

skps23 January 22, 2025, 10:41pm 3

I can get the grid and the block sizes but not the input sizes of the matrices to the kernel via nsight compute. Could you please show it with an example or any link showing this information?

Ncu profiling failed to profile specific kernels

Topic		Replies	Views
Nisght Compute Kernel mapping information Nsight Compute	5	677	October 28, 2022
How to get the compute and memory throughput of GPU from the perspective of the whole GPU system Nsight Compute cuda	4	1245	September 23, 2022
Question about Memory Workload Analysis (keyword) Nsight Compute	4	678	October 12, 2021
Utilization report in Nsight Systems Profiling Linux Targets	2	288	July 4, 2024
transposed matrix of size N*M CUDA Programming and Performance	0	732	August 12, 2011
How to profile a part of kernel function with Nsight Compute Nsight Compute	3	518	April 10, 2024
How do I get the number of shared memory bytes loaded/stored in a kernel Nsight Compute	4	1256	May 15, 2025
How to get memory access profile (over time) inside a kernel? Profiling Linux Targets nsight	4	865	May 3, 2023
How to count memory requests? as reported in nsight analysis CUDA Programming and Performance	0	839	May 31, 2012
NVIDIA Nsight Compute to profile the whole application Nsight Compute	4	604	May 26, 2021

ML workload - getting matrix sizes from profiler

Related topics