Supercomputers/Systems/Pre-Eos
USA · operational · ai training
Pre-Eos
Also known as Pre-Eos 128 Node DGX SuperPOD
Operated by NVIDIA at NVIDIA (undisclosed US site) .
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
- 40.66 PFlop/s
- Rpeak
- 58.04 PFlop/s
- Rmax ÷ Rpeak
- 70.1%
- Cores
- 81,920
- 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 | - NVIDIA DGX H100 128-node DGX SuperPOD | NVIDIA | - | Reported | top500.org |
| Accelerator | NVIDIA H100 SXM5 Accelerators CUDA (Hopper GH100) · 16,896C · 1.98 GHz · TSMC 4N · 700 W NVIDIA H100 Tensor Core GPUs in NVIDIA DGX H100, 128 nodes | NVIDIA | 1,024 estimated | Reported | top500.orgdocs.nvidia.com |
| CPU | Intel Xeon Platinum 8480C CPUs x86-64 · 56C · 2 GHz · Intel 7 · 350 W Xeon Platinum 8480C 56C 2GHz | Intel | - | Reported | top500.org |
| Interconnect | NVIDIA InfiniBand NDR (Quantum-2) Interconnect Fat tree or dragonfly+ · 400 Gb/s per port · Under 0.6 microseconds NVIDIA ConnectX-7 NDR 400G Infiniband | NVIDIA | - | Reported | top500.org |
| Operating system | Ubuntu Server Operating systems Linux distribution · Open source, free with paid support tiers Operating System: Ubuntu 22.04 | Canonical | - | Reported | top500.org |
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 | 1,024 accelerators | 2023-06 | Third-party estimate | Inferred: the TOP500 entry is named 128 Node DGX SuperPOD and NVIDIA's DGX SuperPOD H100 reference architecture puts 8 GPUs in each of 128 nodes (1,024). No source states the count for this entry directly. | top500.orgdocs.nvidia.com |
The read
NVIDIA's 128-node DGX H100 SuperPOD, entered in the June 2023 TOP500 at rank 14 with 40.66 PFlop/s Rmax under the name Pre-Eos, ahead of the larger Eos system that debuted in November 2023. TOP500 lists Xeon Platinum 8480C hosts and NDR 400G InfiniBand. NVIDIA's DGX SuperPOD H100 reference architecture describes a 4-unit, 128-node configuration with 1,024 GPUs, which is how the GPU count here is inferred; neither source states the count for this entry directly. It appeared only in the June 2023 list.
Timeline
Public list appearances
Pointers only: rank and edition, linking to the canonical entry. We do not reproduce list tables. See sourcing policy.
Site & facility
Change history
- 2026-10-09 Added Pre-Eos (NVIDIA)
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 |
|---|---|---|
| docs.nvidia.com | does not name this system | NVIDIA H100 SXM5, NVIDIA InfiniBand NDR (Quantum-2) · counts: 1024 |
| top500.org | names system and part | NVIDIA H100 SXM5, Intel Xeon Platinum 8480C, NVIDIA InfiniBand NDR (Quantum-2), Ubuntu Server |
Checked 2026-10-09 by pnpm verify. Re-run it and the table changes with the web.
Further reading
Inside a node
Why accelerators exist, what the memory hierarchy costs, and how the CPU and accelerator merged onto one package.
The interconnect
Topologies, why latency and tail behaviour matter more than bandwidth, and how the fabric market consolidated.
Programming a supercomputer
The software stack beneath the application, and why the OS is now a sovereignty signal as much as a technical choice.