AI News

Trending story

PyTorch adds compile-time heuristics to Helion kernels

1 reports1 story updatesUpdated Nov 13 00:45

Understand this story

AI overview

On November 13, 2026, the PyTorch blog announced that PyTorch's Helion DSL has added compile-time heuristics, which analyze the kernel IR and data flow to classify it as matmul, reduction, or pointwise, and directly provide configurations such as block size, num_warps, and num_stages without the need for autotuning. According to the blog, on H100, a linear attention kernel is 1.18x faster than the hand-written Triton FLA forward pass, 6 vLLM reduction kernels are 1.27x faster than their hand-written CUDA/C++, and the general reduction kernel suite reaches 1.09x of torch.compile max-autotune. The heuristics have been merged into Helion and are enabled by default on the Triton backend for H100 and B200.

AI-generated from public reports

Latest development · Report

PyTorch Helion Introduces Compile-Time Heuristics, Selecting Kernel Configurations Without Autotuning

Developments

1 confirmed developments

  1. PyTorch adds compile-time heuristics for Helion kernels

    Source:PyTorch Blog

Reports and discussion timeline

Browse reports and discussion by time. Their evidence levels remain distinct.

Nov 13

  1. PyTorch BlogFirst-party

    PyTorch Helion Introduces Compile-Time Heuristics, Selecting Kernel Configurations Without Autotuning

    PyTorch's Helion DSL adds compile-time heuristics that analyze the kernel IR and data flow to classify it as matmul, reduction, or pointwise, and directly provide configurations such as block size, num_warps, and num_stages without autotuning. On H100, the linear attention kernel is 1.18x faster than the hand-written Triton FLA forward pass, vLLM's 6 reduction kernels are 1.27x faster than its hand-written CUDA/C++, and the general reduction kernel suite reaches 1.09x of torch.compile max-autotune. The heuristics have been merged into Helion and are enabled by default on the Triton backend for H100 and B200.

Related stories

  1. Related storiesAnthropic launches Claude Fable 5.2 and Claude Mythos 5.2
  2. Related storiesHao AI Lab integrates MiniMax H3 Flex into FastVideo
  3. Related storiesThe DeepSeek Harness team shares its Super-agent design exploration
  4. Related storiesHao AI Lab open-sources the JetSpec++ model, code, and inference engine