Google TPU v7 (Ironwood)
Google's inference-age flagship: 4.6 PFLOPS FP8 and 9,216-chip superpods.
You're looking at a chip record: what this processor is, its key specs, who designs, manufactures and supplies its parts, and where it gets used.
Ironwood, announced April 2025 and generally available late 2025, is Google's first TPU pitched for the 'age of inference': 4,614 TFLOPS of FP8 per chip, 192 GB of HBM3E at 7.4 TB/s, and pods scaling to 9,216 liquid-cooled chips. Anthropic's October 2025 agreement to access up to one million TPUs — reported at up to a gigawatt of capacity in 2026 — made it the largest publicly known TPU commitment.
Specifications & key facts
Each figure is labeled by basis: vendor spec, vendor claim, reported, estimate, or official. "Not publicly disclosed" means exactly that — we never invent values.
Overview
Memory
Compute
Physical
How this connects to the rest of the stack
This record's links at a glance — click a group to jump to it. +2 more groups below.
Designed by
Memory technology
Cooling
Found inside
Supported by
Used to train
Available from
Successor to
Competes with
Sources
Record reviewed as of 2026-01-15 — see Methodology.