在拉斯维加斯举行的最近一次Google Cloud Next大会上,谷歌发布了专注于推理的新型TPU,其采用名为"Broadfly"的新型网络拓扑结构。 通过采用高基数设计,谷歌可在单个集群中扩展至1,152个TPU。 与Ironwood相比,这使集群规模扩大4.5倍,同时减少网络直径,任意两芯片间最多仅需7次跳转。(1/3) 🧵
During their last Google Cloud Next conference in Las Vegas, Google unveiled their new inference-focused TPU, featuring a novel network topology called "Broadfly". By leveraging a high-radix design, Google can scale up to 1,152 TPUs in a single pod.
Compared to Ironwood, this enables a 4.5x larger pod size while reducing network diameter and with a maximum of just 7 hops between any two chips. (1/3) 🧵