AI chip · Google
Google TPU v6 (Trillium)
Sixth-generation TPU announced May 2024; ~4.7x peak compute over TPU v5e. Backbone of Gemini 2.0-era training runs.
Deployed in 1 named data center · Part of 1 rack design.
Market position
Google's inference-optimised sixth generation (v6e). 4.7x TPU v5e per-chip FP16, 67% more memory bandwidth. Serves Gemini 1.5 Flash and Gemini 2 Nano-class workloads inside Google Cloud. v6p (Ironwood) is the training-optimised sibling.
Efficiency and power
How much work you get per watt and per dollar, and how much power a full rack draws.
- Perf per watt
- 9.18 FP8 TFLOPS/W1,836 TFLOPS ÷ 200 W = 9.18
- Cooling
- waterDepends on the rack design
What fits in 32 GB
Which open-source LLMs run on one Google TPU v6 (Trillium), 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 v6 (Trillium) cost?
Google TPU v6 (Trillium) 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 v6 (Trillium) have?
32 GB of HBM3, running at 1,600 GB/s. Straight from the vendor datasheet. See chips with the most memory for context.
How much power does one Google TPU v6 (Trillium) draw?
200 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 data centers use Google TPU v6 (Trillium)?
1 named data center runs them, including Google Fort Wayne / New Haven. See who has the most Google TPU v6 (Trillium) for the ranked list.
Which open-source LLMs fit on one Google TPU v6 (Trillium)?
Of 13 open-source LLMs we track, 2 fit at FP16, 3 at INT8, and 4 at INT4 (for example Llama 3.1 8B, Qwen 2.5 7B and Mixtral 8x7B). 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
- Most recent source
- 2024-05-01 (across 2 sources on this page)
Specifications (9 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
- 200 W
- Memory
- 32 GB HBM3
- Memory bandwidth
- 1,600 GB/s
- BF16 (dense)
- 918 TFLOPS
- FP8 (dense)
- 1,836 TFLOPS
- INT8 (dense)
- 1,836 TOPS
- Form factor
- OAM (Trillium pod, up to 256 chips)
- Announced
- 2024-05-14
- Released
- 2024-12-01
Source: vendor datasheet
Generation
Foundry & process
- Fabbed at
- TSMC (all chips TSMC makes →)
Rack designs that use it
Whole-rack systems buyers order in bulk (NVL72, HGX, TPU pods) that ship with Google TPU v6 (Trillium) inside.
Compare with
Benchmarks
Published performance numbers by workload. Vendor datasheet figures where noted; independent measurements otherwise.
| Workload | Value | Unit | Source | As of |
|---|---|---|---|---|
| peak tflops bf16(~4.7x TPU v5e) | 918 | TFLOPS | vendor | 2024-05-01 |
Data centers running Google TPU v6 (Trillium)
Who has the most? →Named data-center campuses with Google TPU v6 (Trillium) on site. Counts shown where the operator has published them; other rows are described in general terms.
- Google Fort Wayne / New Haven Data Centeras of 2024-05-01under constructioninferred
Google TPU capacity in Fort Wayne / New Haven campus (Trillium-generation)
Chip mix not published per-campus; inferred from Google's current-generation TPU deployment posture.
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.