How does Triton do Warp Spec?

May 9, 2025

Kapil Sharma from the PyTorch team has a great series of posts diving into the Triton compiler process: 1, 2, 3. As covered there, Triton lowers to a series of intermediate representations, and each level has a set of transformational passes that implement optimizations. TTIR is the generic Triton IR and leverages a number of standard MLIR passes like common subexpression elimination, as well as some Triton specific passes like managing broadcast ops. That’s then lowered to TTGIR, a GPU-specific IR1

  1. Note:Because of some project weirdness, warp specialization is quite different in the release branches from main, so I’ll refer to 3.3 from here on. It’s in very active development (by teams at Meta, OpenAI and Nvidia!) so the specifics are quite likely to change over coming releases! ↩

Read More

Autotuning in PyTorch & Triton

May 4, 2025

torch.compile offers some knobs for controlling the trade-off of execution performance with longer compile times. This is particularly useful for inference, where the same model will be running for a long time.

Read More