Supercomputers/Systems/DESY Maxwell
Germany · operational · mixed
DESY Maxwell
Also known as Maxwell cluster; DESY IDAF Maxwell HPC
Operated by DESY .
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
- 5.60 PFlop/s
- Rmax ÷ Rpeak
- -
- Cores
- 50,328
- 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 |
|---|---|---|---|---|---|
| Interconnect | - InfiniBand fabric with 6 root, 12 top and 42 leaf switches | NVIDIA | - | Reported | docs.desy.de |
| Storage | - GPFS cluster file systems: exfel 60 PB, petra3 30 PB, cssb 11 PB, cfel 1.6 PB | IBM | - | Reported | docs.desy.de |
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 | 447 accelerators | 2026-10 | Operator stated | GPUs across 195 GPU nodes. | docs.desy.de |
The read
Maxwell is the HPC cluster of DESY in Hamburg, serving photon-science data analysis, GPU-accelerated and AI work and general scientific computing. DESY's infrastructure page lists 990 CPU+GPU nodes, 50,328 physical cores (100,656 with hyperthreading), 195 GPU nodes holding 447 GPUs, 600 TB of RAM and a theoretical peak of 5,600 TFlop/s (1,485 TFlop/s CPU plus 4,118 TFlop/s GPU). The InfiniBand fabric has 6 root, 12 top and 42 leaf switches. Cluster file systems include GPFS volumes of 60 PB (exfel), 30 PB (petra3), 11 PB (cssb) and 1.6 PB (cfel) plus 5 PB of DUST. About 4,700 users are recorded. The peak figure is a theoretical, operator-stated number, not an HPL result. Single-source figures from the operator's own documentation, so the tier is reported.
Timeline
No dated events recorded.
Change history
- 2026-10-09 Added DESY Maxwell
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.desy.de | names system only | - |
Checked 2026-10-09 by pnpm verify. Re-run it and the table changes with the web.