Supercomputers/Systems/Tyche-hsg
USA · operational · ai training
Tyche-hsg
Also known as NVIDIA Tyche-hsg GB200 NVL72 cluster
Operated by NVIDIA at NVIDIA (undisclosed US site) .
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 | NVIDIA GB200 NVL72 Accelerators Arm CPU + CUDA GPU · TSMC 4NP 2,048 nodes x 4 NVIDIA Blackwell GB200 GPUs (128x NVIDIA GB200 NVL72 per system name), 8,192 GPUs submitted to MLPerf Training v6.0 | NVIDIA | 8,192 reported | Reported | github.comgithub.com |
| CPU | NVIDIA Grace CPUs Armv9 (Neoverse V2) · 72C · 3.1 GHz · TSMC 4N Two NVIDIA Grace (Neoverse-V2, 72 cores) host processors per node | NVIDIA | - | Reported | github.com |
| Interconnect | NVIDIA InfiniBand NDR (Quantum-2) Interconnect Fat tree or dragonfly+ · 400 Gb/s per port · Under 0.6 microseconds 4x ConnectX-7 InfiniBand NDR 400 Gb/s per node | NVIDIA | - | Reported | github.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 | 8,192 accelerators | 2026-06 | Operator stated | MLCommons v6.0 NVIDIA submission file: 2,048 nodes x 4 GPUs (v5.1: 1,280 nodes, 5,120 GPUs); rack count in the name implies 9,216 installed | github.comgithub.com |
The read
NVIDIA's own GB200 NVL72 cluster as it appears in MLCommons MLPerf Training submissions. The v5.1 file (results of November 2025) names it 'Tyche-hsg (80x NVIDIA GB200 NVL72)' with 1,280 nodes of four GPUs, so 5,120 GPUs in the run; the v6.0 file (repository created 15 June 2026) names it 'Tyche-hsg (128x NVIDIA GB200 NVL72)' with 2,048 nodes, so 8,192 GPUs. In both, only 64 of each rack's 72 GPUs are counted, so the installed total may be higher (128 x 72 = 9,216 by the rack count in the name); the recorded 8,192 is the submitted figure. Listed fabric is four ConnectX-7 InfiniBand NDR 400 Gb/s ports per node with Grace (Neoverse-V2) hosts, and the entry status is 'Available cloud'. NVIDIA's MLPerf Training 6.0 blog reports NVIDIA GB200 NVL72 submissions at 5,120 and 8,192 GPUs but does not use the Tyche-hsg name. No page says where the machine sits and the 'hsg' suffix is unexplained, so the site is NVIDIA's undisclosed one. A smaller 'Tyche (8x NVIDIA GB200 NVL72)' appears in v5.0 and v5.1 and is not counted separately here. It has no HPL result.
Timeline
No dated events recorded.
Site & facility
Change history
- 2026-10-17 Added NVIDIA Tyche-hsg GB200 NVL72 cluster (MLPerf Training v5.1 and v6.0)
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 |
|---|---|---|
| blogs.nvidia.com | does not name this system | NVIDIA GB200 NVL72, NVIDIA InfiniBand NDR (Quantum-2) · counts: 8192 |
| github.com | names system, part and count | NVIDIA GB200 NVL72 · counts: 8192 |
Checked 2026-10-18 by pnpm verify. Re-run it and the table changes with the web.