AI chip · Google
Google TPU v7 (Ironwood)
Seventh-generation TPU announced Cloud Next 2025; positioned as Google's first inference-first TPU generation.
Market position
Google's seventh-generation training TPU (v7p / Ironwood). Announced Apr 2025, GA late 2025. 10x TPU v5p peak throughput. Serves Gemini 3 training. Direct rival to NVIDIA GB200 NVL72 at rack-scale.
Efficiency and power
How much work you get per watt and per dollar, and how much power a full rack draws.
- Perf per watt
- 13.18 FP8 TFLOPS/W9,228 TFLOPS ÷ 700 W = 13.18
What fits in 192 GB
Which open-source LLMs run on one Google TPU v7 (Ironwood), by precision. Weights only: add roughly 15% headroom for real serving. If a model does not fit at FP16, try INT8 or INT4 (smaller quality trade-off than most people expect).
| Model | Params | FP16 | INT8 | INT4 |
|---|---|---|---|---|
| Llama 3.1 8B dense | 8 B | 16 GB ✓ | 8 GB ✓ | 4 GB ✓ |
| Llama 3.1 70B dense | 70 B | 140 GB ✓ | 70 GB ✓ | 35 GB ✓ |
| Llama 3.1 405B dense | 405 B | 810 GB ✗ | 405 GB ✗ | 203 GB ✗ |
| Llama 3.3 70B dense | 70 B | 140 GB ✓ | 70 GB ✓ | 35 GB ✓ |
| DeepSeek V3 MoE (671B total, 37B active per token) | 671 B | 1342 GB ✗ | 671 GB ✗ | 336 GB ✗ |
| DeepSeek R1 MoE (671B total, 37B active per token) | 671 B | 1342 GB ✗ | 671 GB ✗ | 336 GB ✗ |
| Qwen 2.5 7B dense | 7 B | 14 GB ✓ | 7 GB ✓ | 4 GB ✓ |
| Qwen 2.5 72B dense | 72 B | 144 GB ✓ | 72 GB ✓ | 36 GB ✓ |
| Mixtral 8x7B MoE (46.7B total, 12.9B active per token) | 46.7 B | 93 GB ✓ | 47 GB ✓ | 23 GB ✓ |
| Mixtral 8x22B MoE (141B total, 39B active per token) | 141 B | 282 GB ✗ | 141 GB ✓ | 71 GB ✓ |
| Gemma 2 27B dense | 27 B | 54 GB ✓ | 27 GB ✓ | 14 GB ✓ |
| Command R+ dense | 104 B | 208 GB ✗ | 104 GB ✓ | 52 GB ✓ |
| Kimi K2 MoE (1T total, 32B active per token) | 1000 B | 2000 GB ✗ | 1000 GB ✗ | 500 GB ✗ |
Math: FP16 = params × 2 bytes; INT8 = params × 1 byte; INT4 = params × 0.5 bytes. MoE models sum every expert (full weights on disk), not the per-token active subset.
Common questions
How much does Google TPU v7 (Ironwood) cost?
Google TPU v7 (Ironwood) doesn't have a public launch list price. Vendors like this one usually price through direct sales rather than a public sheet. See list price per TFLOP for chips that do publish.
How much memory does Google TPU v7 (Ironwood) have?
192 GB of HBM3e, running at 7,400 GB/s. Straight from the vendor datasheet. See chips with the most memory for context.
How much power does one Google TPU v7 (Ironwood) draw?
700 W at the chip. A full server draws more once you add CPU, memory, networking and cooling: see the rack power number above. Compare to other chips on perf-per-watt.
Which open-source LLMs fit on one Google TPU v7 (Ironwood)?
Of 13 open-source LLMs we track, 7 fit at FP16, 9 at INT8, and 9 at INT4 (for example Llama 3.1 8B, Llama 3.1 70B and Llama 3.3 70B). Full table above with each model's memory need.
Key facts
- Class
- Compute SoC (AI accelerator)
- Designer
- Safety-critical?
- No (data-center inference / training)
- Record as of
- 2026-09-27
Specifications (8 fields, click to expand)
Straight from the vendor datasheet. Dense throughput shown first; sparse (2:4) numbers in parentheses where the vendor publishes them. Full datasheet linked below.
- TDP
- 700 W
- Memory
- 192 GB HBM3e
- Memory bandwidth
- 7,400 GB/s
- BF16 (dense)
- 4,614 TFLOPS
- FP8 (dense)
- 9,228 TFLOPS
- Form factor
- OAM (Ironwood pod, up to 9,216 chips)
- Announced
- 2025-04-09
- Released
- 2025-12-01
Source: vendor datasheet
Generation
Foundry & process
- Fabbed at
- TSMC (all chips TSMC makes →)
Compare with
Sources
Adoption rows appear as we confirm each chip-in-robot pairing from a public source. See every chip we track for the full catalog or the Google page.