https://github.com/probcomp/warped

nvidia warp profiling
https://github.com/probcomp/warped

Last synced: about 1 month ago
JSON representation

nvidia warp profiling

• Host: GitHub
• URL: https://github.com/probcomp/warped
• Owner: probcomp
• Created: 2024-10-28T23:16:06.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2024-11-07T17:48:44.000Z (over 1 year ago)
• Last Synced: 2026-02-05T11:52:37.107Z (about 2 months ago)
• Size: 3.58 MB
• Stars: 0
• Watchers: 4
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# Examples

We can run and profile all the [examples](https://github.com/NVIDIA/warp/tree/main/warp/examples) in Warp. Let's go through the raycast bunny example.

## Raycast Bunny



Profile the [raycast bunny](https://github.com/NVIDIA/warp/blob/main/warp/examples/core/example_raycast.py):

```shell

sudo ~/.pixi/bin/pixi run ncu \

  --set full \

  --target-processes all \

  --metrics dram__bytes_read.sum,dram__bytes_written.sum,sm__inst_executed_pipe_tensor_op_hmma.avg,sm__cycles_elapsed.avg,l2_tex_read_bytes.sum,l2_tex_write_bytes.sum,lts__t_bytes.sum,lts__t_sectors_pipe_lsu_mem_rd.sum,lts__t_sectors_pipe_lsu_mem_wr.sum  \

  --nvtx \

  --call-stack \

  --export example_raycast_bunny \

  python -m warp.examples.core.example_raycast

```

Then open the `example_raycast_bunny.ncu-rep` report from Nsight Compute UI.