{"data":{"slug":"google-tpu-v5p","etype":"chip","name":"Google TPU v5p","cat":"AI Accelerator","subtitle":"The training TPU behind Gemini 1.0 — 8,960-chip pods.","summary":"Google's performance TPU of the Gemini era: 459 BF16 TFLOPS and 95 GB of HBM per chip, scaled into pods of 8,960 chips over its inter-chip interconnect. Gemini 1.0 was trained on TPU v4 and v5p fleets — evidence that frontier models could be built entirely off NVIDIA hardware.","attrs":[{"g":"Overview","l":"Designer","v":"Google"},{"g":"Overview","l":"Announced","v":"December 2023"},{"b":"Vendor spec","g":"Memory","l":"Memory","v":"95 GB HBM2e"},{"b":"Vendor spec","g":"Memory","l":"Memory bandwidth","u":"TB/s","v":"2.77"},{"b":"Vendor spec","g":"Compute","l":"BF16","u":"TFLOPS","v":"459"},{"b":"Vendor spec","g":"Compute","l":"INT8","u":"TOPS","v":"918"},{"b":"Vendor spec","g":"Physical","l":"Pod scale","v":"8,960 chips; ICI 4.8 Tb/s per chip"}],"kf":{"year":2023,"class":"AI Accelerator","mem_bw":2.77,"memory":"95 GB HBM2e","designer":"Google"},"aliases":["TPU v5p"],"ticker":null,"exchange":null,"hq":null,"founded":null,"website":null,"status":"Shipping","region":null,"asof":"2026-01-15","sources":[{"url":"https://cloud.google.com/tpu/docs/v5p","label":"Google Cloud TPU v5p"}]},"meta":{"found":true,"count":1,"served_from":"site cache (no live upstream calls)","as_of":"2026-10-14T07:56:03+00:00","disclaimer":"Research and information only — not investment advice. Structured records carry basis labels (vendor spec / reported / estimate); market quotes are delayed/indicative.","attribution":"Records compiled from vendor documentation, official filings and reputable public reporting; market quotes via financial data providers. See /data-sources."}}