Using Nsight Compute to profile kernels in vLLM without creating repro scripts
(blog.ncompass.tech)
1 pts|27 days ago|discuss
user: adiraja
19 karma | created 2 years ago
recent submissions
1 month ago|discuss
Runtime injection of temporary debug code
(loom.com)
1 pts|1 month ago|1 comment
A VSCode extension for GPU/AI trace profiling, viewing and performance analysis
(marketplace.visualstudio.com)
3 pts|3 months ago|2 comments
1 year ago|discuss
1 year ago|discuss
1 year ago|discuss
1 year ago|discuss
1 year ago|discuss
1 year ago|discuss
1 year ago|discuss
1 year ago|discuss
37 pts|1 year ago|34 comments