Google open-sources TorchTPU with Meta, letting PyTorch workloads run natively on TPUs
mark_k · x · 2026-10-21
Google, built with Meta's PyTorch team, has open-sourced TorchTPU, letting existing PyTorch workloads run natively on Google TPUs with minimal code changes across training and inference. It integrates with Hugging Face, vLLM and SGLang; RadixArk says its SGLang integration already runs at large-scale production in Meta, with the same serving code working across GPUs and TPUs including prefix caching and speculative decoding. Meta's PyTorch engineering director: "We cannot be constrained by a single hardware architecture."
More from Infra
- 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
- Ollama Runs Open Models on NVIDIA Blackwell via Crusoe, Live in Two Weeks — ollama · 2026-10-22
- CUDA's moat quietly eroding as Codex can rewrite kernels for AMD, investor notes — paigeinsf · 2026-10-22
- Claim: every 2nd/3rd-tier inference provider now has a recipe for 90%+ compute margins — zephyr_z9 · 2026-10-22