Google TPU v7 Ironwood

Owner The company that specified the chip and runs it in its own cloud, data centers or products.

Google 1

Role What the owner says the chip is for: training, inference, both, or a general-purpose server CPU.

Training and inference 1

Design partner The ASIC house that co-designed or physically implemented the chip (Broadcom, Marvell, MediaTek, Alchip, GUC and others), as stated by the owner, the partner, a filing or named trade-press reporting. 'In-house' when the owner states it designed the chip itself, including through an acquired team such as Annapurna Labs.

Broadcom 4

Partner basis Who says so: the owner, the partner, or press reporting with the outlet named. A reported partner is not confirmed by the owner.

Reported by CNBC: Broadcom's CEO said Broadcom sells Ironwood TPU racks to Anthropic. Google has not named a design partner. 4

Memory capacity GB On-package memory per chip in GB. Blank for CPUs, whose memory sits off-package.

192 1

Memory bandwidth TB/s Peak memory bandwidth per chip in TB/s, as published.

7.37 1

BF16 TFLOPS Published BF16 throughput per chip or accelerator module in TFLOPS. FP16 is a separate column. See the precision note for sparsity and module scope.

2,307 2

FP8 TFLOPS Published peak FP8 throughput per chip in TFLOPS. The precision note records sparsity where the owner states it. Blank when the owner gives no FP8 figure.

4,614 1

Precision note What the throughput figures mean: dense or sparse, which precision, per chip or per package, and any figure the owner states only in another unit.

Per-chip figures: Google states 4,614 TFLOPS FP8 and 2,307 TFLOPS BF16 and does not label them dense or sparse. Two chiplets with 96 GiB HBM each. 2

Largest pod chips The largest scale-up domain the owner states: chips per pod, UltraServer, supernode or rack joined by the chip-to-chip interconnect.

9,216 1

Status Announced: public roadmap without verified deployment. Sampling: first silicon or limited internal use. Preview: customers can request early access, without a GA claim. Deployed: stated production use or general availability. Discontinued: program stopped.

Deployed 2

Announced Year the owner first announced the chip publicly.

2,025 1

Deployment scale The largest deployment the owner or a named report states: chip counts, clusters, gigawatts or regions.

Anthropic plans to access up to 1 million TPUs (Google, Nov 2025); Broadcom took a $10B order to sell Ironwood racks to Anthropic (Broadcom CEO, Dec 2025) 3

How to get it How an outside buyer can use the chip: a named cloud instance or service, a product, or 'Internal only'.

Google Cloud TPU7x (Ironwood) 2

Not verified in the sources reviewed: foundry, node, memory, fp16, mxfp8, mxfp4, mx8, mx4, int8, fp4, power.

1. Google: Ironwood: The first Google TPU for the age of inference accessed 2026-11-11
2. Google Cloud: TPU7x (Ironwood) accessed 2026-11-11
3. Google Cloud Blog: Ironwood TPUs and new Axion-based VMs for your AI workloads accessed 2026-11-11
4. CNBC: Broadcom reveals its mystery $10 billion customer is Anthropic accessed 2026-11-11