Google I/O

Accelerate PyTorch workloads with Cloud TPUs and OpenXLA

Google Cloud AI accelerators enable high-performance, cost-effective training and inference for leading AI/ML frameworks. In this session, get the latest news on PyTorch/XLA, developed collaboratively by Google, Meta, and partners in the AI ecosystem to utilize OpenXLA for accelerating PyTorch workloads.

Advanced
Technical session
Join a community group

Meet developers, discover local groups, and build your global network.

Get started
I/O Connect

Explore, network, and get hands-on with the latest products.