Meta engineer confirms Google TPU inference is running in production

What happened

Baoyu reports that Meta’s PyTorch-TPU lead, Zongwei Zhou, confirmed SGLang-TorchTPU is running in Meta’s production environment, citing Google’s October 20 developer blog announcing SGLang inference on TPU. Zhou did not disclose deployment scale, affected products, or whether the TPUs are rented or purchased. The Information reported in February that Meta and Google signed a multiyear TPU rental contract worth billions of dollars; neither company publicly detailed the arrangement.

The story so far

Meta’s PyTorch-TPU lead, Zongwei Zhou, confirmed that SGLang-TorchTPU is running in Meta’s production environment for AI inference, according to Baoyu. The confirmation follows Google’s announcement of SGLang inference support on TPU. The Information reported in February that Meta and Google had signed a multiyear TPU rental contract worth billions of dollars. Zhou’s confirmation establishes production use but does not establish whether that deployment uses rented or purchased TPUs.

Keep reading

Trending now

1. 1Mistral Releases Voxtral 3 Speech Models, Voice Agents API and LiveKit-Based SDK
2. 2DeepSeek Open-Sources DeepInfer and API Components, With Model and Hardware Limits
3. 3OpenAI Acquires Orchestra Bio to Expand AI Tools for Drug Development
4. 4Google Rolls Out Gemini Personal Intelligence to Android Auto, With Google Built-In Cars to Follow