Supercomputers/Systems/Google TPU v6e (Trillium) pod
USA · operational · mixed
Google TPU v6e (Trillium) pod
Also known as Trillium-256 pod
Operated by Alphabet at Google Cloud (undisclosed location) .
Researched with AI assistance, checked by scripts. These rows were researched by AI agents reading public operator, laboratory and vendor sources, appended by scripts and checked by automated validation; no subject-matter expert has reviewed them (how this site is made). A mechanical second-reader pass has since fetched every cited source: 1350 of 1540 systems have a readable citation that names them, and 64 are genuinely weakly sourced. Every claim carries its source and a confidence tier. Treat anything below verified as a lead, not a citation.
Measured performance
- Rmax
- -
- Rpeak
- -
- Rmax ÷ Rpeak
- -
- Cores
- -
- Power
- -
- Per watt
- - GF/W
Figures are as last publicly reported for the configuration described below, not a live measurement. Where a system was upgraded in place, the post-upgrade configuration is shown and the earlier one appears in the timeline.
What this machine is made of
One row per supplier relationship. “Supplier at build” is the company that shipped the part at the time; where that company has since been acquired, the parent it rolls up to today is shown beside it. That distinction is what makes ticker-level aggregation possible across a thirty-year dataset.
| Role | Part | Supplier at build | Quantity | Confidence | Source |
|---|---|---|---|---|---|
| Accelerator | Google TPU v6e (Trillium) Accelerators TPU (systolic array) TPU v6e (Trillium): 256 chips in a pod, 16x16 topology on 32 hosts | Alphabet | 256 reported | Reported | docs.cloud.google.comcloud.google.com |
| Interconnect | - TPU inter-chip interconnect, 2D torus, 4 ICI ports per chip at 800 GBps bidirectional | Alphabet | - | Reported | docs.cloud.google.com |
AI datacenter metrics
Figures beyond an accelerator count, each with who stated it. What can and cannot be compared across AI datacenters.
| Metric | Value | As of | Basis | Note | Source |
|---|---|---|---|---|---|
| Accelerators | 256 accelerators | 2026-10 | Operator stated | Google Cloud docs: 256 chips per v6e pod; one documented pod configuration, not a site | docs.cloud.google.comcloud.google.com |
The read
One Cloud TPU v6e (Trillium) pod as Google documents it: 256 chips in a 16x16 2D torus on 32 hosts of eight chips (1,536 GiB of DRAM and four 200 Gbps NICs per host), each chip 918 TFLOPs bf16 with 32 GB of HBM at 1,638 GBps and 800 GBps of bidirectional inter-chip bandwidth. Google's own per-pod peak is 234.9 PFLOPs bf16 (a bf16 peak, not an HPL figure, so no FLOPS is recorded). Google's December 2024 general-availability post adds that pods join over its Jupiter data-center network, up to 100,000 Trillium chips in one fabric, and reports 99% scaling efficiency at 12 pods (3,072 chips) and 94% at 24 pods (6,144 chips) on GPT-3 175B. Like google-tpu-v5p-pod this is a documented product configuration, not a located installation: neither page says where pods sit or how many exist, so USA is the operator's home country and the site is Google's undisclosed one. Per-chip bf16 is about 4.7x the v5e (google-tpu-v5e-pod). Epoch AI's modelled campuses google-columbus-tpu, google-new-albany-tpu and google-council-bluffs-east-tpu are the located v6e capacity.
Site & facility
Change history
- 2026-10-17 Added Google TPU v6e (Trillium) 256-chip pod (product configuration)
Source check
We fetched this system's own citations and recorded whether each page actually mentions it. This is a corroboration signal, not a fact check, and it is published so you can see how well the sourcing holds up rather than take it on trust.
| Cited source | Result | Found on the page |
|---|---|---|
| cloud.google.com | names system, part and count | Google TPU v6e (Trillium) · counts: 256 |
| docs.cloud.google.com | names system, part and count | Google TPU v6e (Trillium) · counts: 256 |
Checked 2026-10-18 by pnpm verify. Re-run it and the table changes with the web.