Google and Meta open-source TorchTPU, bringing native PyTorch support to TPUs
PyTorch · x · 2026-10-22
Google and Meta have open-sourced TorchTPU, a native PyTorch backend for Google TPUs that lets existing PyTorch workloads run out of the box.
- Started at Google in mid-2025 in collaboration with Meta's PyTorch team; announced at PyTorch Conf 2026
- Unlike the older PyTorch/XLA (Lazy Tensors, intrusive code changes, high compile latency), TorchTPU prioritizes eager execution and integrates torch.compile and native hardware acceleration
- Works with vLLM, SGLang, TorchTitan, Hugging Face, and Ray
A significant step in opening the TPU ecosystem to the mainstream PyTorch community.
More from Infra
- FlyDSL Kernels Land in PyTorch, Up to 3.01x vLLM Speedups on AMD MI350 — PyTorch · 2026-10-22
- Micron AI memory roadmap: bandwidth moves onto the die, HBM supply tight from 2029 — BenBajarin · 2026-10-22
- Qualcomm's non-phone revenue seen growing 60% next year, driven by auto and embedded — pdamodaran · 2026-10-22
- After PHP advisories, a WebAssembly runtime vendor lays out its security validation plan — MickeySteamboat · 2026-10-22
- Hugging Face: Specializing an SLM's vocabulary can cut inference cost by 77% — Hugging Face Blog · 2026-10-22
- Ollama Runs Open Models on NVIDIA Blackwell via Crusoe, Live in Two Weeks — ollama · 2026-10-22