Supercomputers/Systems/Nabuchodonosor
France · operational · ai training
Nabuchodonosor
Also known as Scaleway Nabu 2023 AI supercomputer
Operated by Scaleway at Scaleway DC5 (Paris) .
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 |
|---|---|---|---|---|---|
| Accelerator | NVIDIA H100 SXM5 Accelerators CUDA (Hopper GH100) · 16,896C · 1.98 GHz · TSMC 4N · 700 W 127 NVIDIA DGX H100 systems with 1,016 H100 Hopper GPUs in a DGX SuperPOD | NVIDIA | 1,016 verified | Verified | scaleway.comnextplatform.com |
| Interconnect | NVIDIA InfiniBand NDR (Quantum-2) Interconnect Fat tree or dragonfly+ · 400 Gb/s per port · Under 0.6 microseconds NVIDIA Quantum-2 InfiniBand at 400 Gb/s in a non-blocking full fat tree, plus NVLink within each DGX | NVIDIA | - | Verified | scaleway.comnextplatform.com |
| Storage | - DDN A3I storage cluster, 1.8 PB, 2.7 TB/s read and 1.95 TB/s write across the DGX nodes | DDN | - | Reported | nextplatform.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 | 1,016 accelerators | 2023-10 | Operator stated | 127 DGX H100 systems, 1,016 H100 GPUs, per Scaleway. | scaleway.com |
| PUE | 1.25 | 2024-12 | Operator stated | Scaleway's 2024 PUE for DC5 (PAR2), the hosting data centre. Its custom-clusters page quotes 1.16 for the same site; the two do not agree. | scaleway.comscaleway.com |
| WUE | 0.25 L/kWh | 2024-12 | Operator stated | Scaleway's 2024 WUE for DC5 (direct free cooling with adiabatic assist), the hosting data centre. | scaleway.com |
The read
Scaleway's first AI supercomputer, nicknamed Nabuchodonosor (a 15-litre champagne bottle), bought by Scaleway's telecom parent iliad Group and installed in the fall of 2023 in the Paris-based DC5 data center. An NVIDIA DGX SuperPOD of 127 DGX H100 systems with 1,016 H100 GPUs, linked by 400 Gb/s NVIDIA Quantum-2 InfiniBand in a non-blocking fat tree, with a DDN A3I storage cluster of 1.8 PB (Next Platform). Scaleway sells it as GPU cloud, with a 16-GPU Jeroboam variant launched alongside. Next Platform cites a vendor peak of over 4 exaflops at FP8 tensor precision, a low-precision AI rate that is not an HPL result, so no Rmax or Rpeak is recorded. DC5 is described as using adiabatic and outside-air cooling. Exact street location and power draw are not published.
Site & facility
Scaleway DC5 (Paris)
- Location
- Paris, France
- Cooling
- Adiabatic and outside-air cooling
- Commissioned
- 2023
Change history
- 2026-10-04 Added AI datacenter metrics for nabuchodonosor
- 2026-10-02 Added Nabuchodonosor, Scaleway's 1,016-GPU H100 DGX SuperPOD
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 |
|---|---|---|
| nextplatform.com | names system, part and count | NVIDIA H100 SXM5, NVIDIA InfiniBand NDR (Quantum-2) · counts: 1016 |
| scaleway.com | names system, part and count | NVIDIA H100 SXM5, NVIDIA InfiniBand NDR (Quantum-2) · counts: 1016 |
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.
Storage and I/O
Parallel file systems, why checkpointing dominates the write load, and the metadata failure mode nobody expects.