PyTorch Adds Mark Kernels Context Manager for CUDA Graph Profiling
——0——
QUOTE POST
#819Edward Z. Yang@EZYANG
This is not just nightlies, it is in 2.12! (Thanks Natalia for the correction)
Pro-tip: using CUDA graphs and annoyed that all the kernels have no labels in your profiles? Get a nightly that has mark_kernels context manager: https://github.com/pytorch/pytorch/pull/179768 (thanks Natalia and Shangdi for implementing!) You need 13.1 driver, but user mode driver is enough
3:07 AM · May 22, 2026 · 6.5K Views
4:54 PM · May 22, 2026 · 2.2K Views