Supercomputers/Systems/Raven-GPU
Germany · operational · mixed
Raven-GPU
Also known as Raven
Operated by MPCDF at MPCDF Garching .
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
- 8.62 PFlop/s
- Rpeak
- 16.03 PFlop/s
- Rmax ÷ Rpeak
- 53.8%
- Cores
- 96,768
- Power
- 377 kW
- Per watt
- 22.9 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 | - Lenovo ThinkSystem SD650-N V2 | Lenovo | - | Verified | top500.orgmpcdf.mpg.de |
| Accelerator | NVIDIA A100 SXM4 40GB Accelerators CUDA (Ampere GA100) · 6,912C · 1.41 GHz · 7 nm · 400 W 192 GPU nodes x 4 NVIDIA A100 40GB SXM GPUs, NVLink3 | NVIDIA | 768 reported | Reported | mpcdf.mpg.de |
| CPU | - Intel Xeon Platinum 8360Y 36C 2.4GHz | Intel | - | Verified | top500.orgmpcdf.mpg.de |
| Interconnect | NVIDIA InfiniBand HDR Interconnect Fat tree or dragonfly+ · 200 Gb/s per port · About 0.6 microseconds Mellanox HDR InfiniBand, 200 Gbit/s to GPU nodes | NVIDIA | - | Verified | mpcdf.mpg.detop500.org |
The read
The GPU partition of Raven, the Max Planck Computing and Data Facility's flagship system at Garching, serving the Max Planck Society's roughly 80 research institutes. A Lenovo ThinkSystem SD650-N V2 cluster of Intel Xeon Platinum 8360Y hosts each paired with four NVIDIA A100 40GB SXM GPUs over NVLink, connected by 200 Gbit/s Mellanox HDR InfiniBand. TOP500 lists Rmax 8.62 PFlop/s, matching MPCDF's own published 8.6 PFlop/s figure, so the tier is verified. First appeared on the June 2021 TOP500 list at rank 47. A separate CPU-only partition, sharing the Raven name but a distinct TOP500 entry (1,592 Xeon Ice Lake-SP nodes, no GPUs), is not recorded as its own row here.
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
- 2026-09-24 Raven-GPU added, the Max Planck Society's GPU-accelerated flagship system at Garching
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 |
|---|---|---|
| top500.org | names system, part and count | NVIDIA A100 SXM4 40GB, NVIDIA InfiniBand HDR · counts: 768 |
| mpcdf.mpg.de | names system and part | NVIDIA A100 SXM4 40GB, NVIDIA InfiniBand HDR |
Checked 2026-10-09 by pnpm verify. Re-run it and the table changes with the web.
Further reading
Operating a supercomputer
Acceptance testing, why "installed" and "in production" are six months apart, and the machine lifecycle.
Inside a node
Latency-optimised versus throughput-optimised processors, NUMA, and the end of the host-to-device copy.
The interconnect
Topologies, why latency and tail behaviour matter more than bandwidth, and how the fabric market consolidated.