/
← Voltar ao feed
low

Google open-sources TorchTPU, joining Amazon's TorchNeuron in challenging Nvidia's hold on PyTorch AI workloads

1 fonte(s) · 1 atualização(ões) · publicado 20/10/2026 15:38 · coletado há 10 horas

On Tuesday Google released TorchTPU as an open-source project. It lets PyTorch AI applications run on Google's Tensor Processing Units (TPUs) and Google Cloud through a configuration change, with no application rewrite. A day earlier Amazon announced TorchNeuron, which ports PyTorch apps to its Trainium chip and AWS. Amazon is also submitting code upstream to decouple PyTorch features from Nvidia hardware. Google says it worked with Meta, Hugging Face, Inferact (vLLM) and RadixArk (SGLang) on compatibility. It also launched a TPU agent suite to help migrate apps and fix errors, and updated LiteRT to work better with PyTorch. This is industry and product news with no security threat, vulnerability or incident. The article cautions that developers are unlikely to abandon Nvidia hardware en masse.

Linha do tempo

  1. techcrunch.com ↗ · pub. 20/10/2026 15:38 · coletado 20/10 16:00 primeiro registro

    Google open-sources TorchTPU, joining Amazon's TorchNeuron in challenging Nvidia's hold on PyTorch AI workloads

    ler a análise desta fonte

    On Tuesday Google released TorchTPU as an open-source project. It lets PyTorch AI applications run on Google's Tensor Processing Units (TPUs) and Google Cloud through a configuration change, with no application rewrite. A day earlier Amazon announced TorchNeuron, which ports PyTorch apps to its Trainium chip and AWS. Amazon is also submitting code upstream to decouple PyTorch features from Nvidia hardware. Google says it worked with Meta, Hugging Face, Inferact (vLLM) and RadixArk (SGLang) on compatibility. It also launched a TPU agent suite to help migrate apps and fix errors, and updated LiteRT to work better with PyTorch. This is industry and product news with no security threat, vulnerability or incident. The article cautions that developers are unlikely to abandon Nvidia hardware en masse.