Triton (OpenAI)
Software
Compute Platform
Active
Python-level GPU kernels — the community's crowbar under the CUDA moat.
You're looking at a software record: a tool or platform in the AI stack — who maintains it and what hardware it runs on.
Triton lets researchers write high-performance GPU kernels in Python instead of raw CUDA C++, compiling to competitive code on NVIDIA and, increasingly, AMD hardware. PyTorch's compiler emits Triton, making it the neutral kernel layer much of the ecosystem now rides on.
How this connects to the rest of the stack
This record's links at a glance — click a group to jump to it.
Runs on
Maintained by
Related
Sources
Record reviewed as of 2026-01-15 — see Methodology.