TensorWave Tucson MI325X Cluster

Operated by TensorWave at TensorWave Tucson Data Center .

Phase 1 seed dataset, compiled by hand. These rows were built from public operator, laboratory and vendor sources. A mechanical second-reader pass has since fetched every cited source: 1092 of 1153 systems have a readable citation that names them, and 40 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
Integrator	- TensorWave-built and operated AMD-only GPU cloud cluster in Tucson, Arizona	TensorWave	-	Reported	siliconangle.com
Accelerator	AMD Instinct MI325X Accelerators ROCm (CDNA 3) · TSMC N5/N6 · 1,000 W 8,192 AMD Instinct MI325X accelerators (Tom's Hardware, Futuriom)	AMD AMDNASDAQ	8,192 reported	Verified	tomshardware.com futuriom.com
Cooling	Direct-to-chip cold plate Cooling Cold plate on the package, air for the remainder · Removes roughly 50-80 percent of node heat to liquid Direct-to-chip liquid cooling loop with cold plates mounted on each MI325X	TensorWave	-	Reported	tomshardware.com

The read

TensorWave's 8,192-GPU AMD Instinct MI325X training cluster, described by the company as the largest AMD training deployment in North America and the neocloud row in this dataset that is built entirely on AMD accelerators. Tom's Hardware (July 2025) and Futuriom (August 2025) independently give the 8,192 MI325X count and describe the cluster as deployed and running, with Tom's Hardware also describing a direct-to-chip liquid cooling loop; SiliconANGLE's May 2025 funding report places it at a data center campus in Tucson, Arizona, where the cluster was then targeted for completion by the end of June 2025 with about 1,000 GPUs installed. Aggregate compute is stated by the trade press as roughly 21 exaflops of FP8, a vendor peak figure that is not recorded as FLOPS here. No source states the server vendor, network fabric or a site power figure for this specific cluster. TensorWave later announced a further 20 MW split between Tucson and Pennsylvania with TECfusions. It has never submitted an HPL result.

Timeline

2025-05 Announced source
2025-07 First operational source

Site & facility

TensorWave Tucson Data Center

Location: Tucson, AZ, USA
Cooling: Direct-to-chip liquid cooling
Commissioned: 2025

siliconangle.com prnewswire.com

Change history

2026-10-01 Added TensorWave Tucson MI325X Cluster (TensorWave)

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
futuriom.com	names system only	· counts: 8192
tomshardware.com	names system, part and count	AMD Instinct MI325X, Direct-to-chip cold plate · counts: 8192

Checked 2026-10-09 by pnpm verify. Re-run it and the table changes with the web.