Google open-sources TorchTPU, its PyTorch port for TPU chips

On October 20, 2026, Google announced that TorchTPU, its long-awaited version of PyTorch, is now available as an open-source project. The release allows AI applications to run on Google's in-house Tensor Processing Units (TPUs) and enables developers to migrate existing PyTorch apps to TPUs and Google Cloud through a simple configuration change, without rewriting their applications.

TorchTPU is Google's adaptation of PyTorch, the open-source AI framework originally developed by Meta in 2016 and transferred to the Linux Foundation in 2022. Google said it worked with Meta, Hugging Face, Inferact, maker of vLLM, and RadixArk, maker of SGLang, to ensure TorchTPU's features and code libraries work well with these popular AI development tools.

Google also launched a dedicated TPU agent suite to help developers move their apps to Google's chips and cloud and fix errors along the way. It also updated LiteRT, its tool for running AI on phones, laptops and embedded devices, to work better with PyTorch AI apps.

The move follows Amazon's announcement on October 19, 2026, of TorchNeuron, its own version of PyTorch that lets AI apps be ported to Amazon's Trainium chip and AWS without rewriting. Amazon is also submitting code to the main PyTorch project to decouple features from Nvidia's chips. Both companies are working to reduce Nvidia's hold on the AI chip market.