技术档案

TPUTensor Processing Unit

AI专用计算加速器

每条都摘自下方编号的来源原文；只写事实，不写预测或建议。

技术简介

TPU是Google定制开发、用于加速机器学习负载的专用集成电路。5
Tensor Processing Units (TPUs) are Google's custom-developed, application-specific integrated circuits (ASICs) used to accelerate machine learning workloads.
TPU已成为Nvidia GPU之外处理AI任务的一种替代选择。6
Sin embargo, los chips pronto demostraron su utilidad para gestionar tareas de inteligencia artificial y, en los últimos años, se han consolidado como una alternativa a las muy apreciadas unidades de procesamiento gráfico de Nvidia.

Anthropic使用Google Cloud的BigQuery、Cloud Run及AlloyDB等支持其数据、AI开发及应用。2
Anthropic also continues to grow its use of Google Cloud's broader cloud and AI solutions, including BigQuery, Cloud Run, AlloyDB, and others, which together help power Anthropic's data, AI development, and applications.
Anthropic使用Google TPU、AWS Trainium及NVIDIA GPU训练和运行Claude。3
We train and run Claude on a range of AI hardware—AWS Trainium, Google TPUs, and NVIDIA GPUs—which means we can match workloads to the chips best suited for them.
Google的Gemini及Anthropic的Claude等模型已用TPU训练和服务。4
Today’s frontier models, including Google’s Gemini, Veo, Imagen, and Anthropic’s Claude train and serve o n Tensor Processing Units (TPUs).

Google在2026年第三季度财报称TPU系统销售对收入增长有实质贡献。7
TPU systems sales were also a meaningful contributor to revenue growth.
博通2026年4月披露Anthropic承诺从2027年起通过博通获取约3.5吉瓦下一代TPU算力。1
Separately, Broadcom, Google and Anthropic PBC (“Anthropic”) have expanded their current strategic collaboration under which Anthropic, beginning in 2027, will access through Broadcom approximately 3.5 gigawatts as part of the multiple gigawatts of next generation TPU-based AI compute capacity committed by Anthropic.

博通与Google签约开发并供应后续代际的定制TPU，并供应AI机柜网络等组件。1
Broadcom Inc. (“Broadcom”) and Google LLC (“Google”) have entered into a Long Term Agreement for Broadcom to develop and supply custom Tensor Processing Units (“TPUs”) for Google’s future generations of TPUs and a Supply Assurance Agreement for Broadcom to supply networking and other components to be used in Google’s next-generation AI racks through up to 2031.

Anthropic消费扩展算力的前提是其持续商业成功。1
The consumption of such expanded AI compute capacity by Anthropic is dependent on Anthropic’s continued commercial success.