PyTorchCon 2026 Poster to Show Warm-Start Autotuning for Helion GPU Kernel DSL

PyTorch · x · 2026-09-24

At PyTorchCon North America 2026 (Oct 20-21, San Jose), Red Hat senior engineer Alessandro Sangiorgi will present a poster on warm starts for Helion autotuning, showing how cached configurations let developers iterate faster on GPU kernels instead of repeating cold searches. Helion is a PyTorch DSL for GPU kernels.

Related event: PyTorchCon to Showcase Helion Autotuning Warm Starts(2 posts)→

Original post →

More from Infra

Infra channel →