Supercomputers/Systems/Dardel
Sweden · operational · mixed
Dardel
Operated by PDC Center for HPC at PDC Data Centre, KTH .
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
- 9.90 PFlop/s
- Rpeak
- 10.94 PFlop/s
- Rmax ÷ Rpeak
- 90.5%
- Cores
- 56,640
- Power
- 167.71 kW
- Per watt
- 59.0 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 | - HPE Cray EX (CPU and GPU partitions) | Hewlett Packard Enterprise | - | Reported | pdc.kth.setop500.org |
| Accelerator | AMD Instinct MI250X Accelerators ROCm (CDNA 2) · 1.7 GHz · TSMC N6 · 560 W GPU partition: 62 nodes, four AMD Instinct MI250X (eight GCDs) per node | AMD | - | Reported | pdc.kth.setop500.org |
| CPU | AMD EPYC 7742 (Rome) CPUs x86-64 · 64C · 2.25 GHz · 7 nm · 225 W CPU partition: 1,278 nodes, 163,584 cores of AMD EPYC Zen 2 (EPYC 7742 per TOP500) | AMD | - | Reported | pdc.kth.setop500.org |
| CPU | AMD EPYC 7A53 (Trento) CPUs x86-64 · 64C · 2 GHz · 7 nm · 280 W GPU nodes: one 64-core AMD EPYC (3rd Gen optimized, 2 GHz) per node | AMD | - | Reported | top500.orgpdc.kth.se |
| Interconnect | HPE Slingshot-11 Interconnect Dragonfly · 200 Gb/s per port · Around 2 microseconds HPE Slingshot-11, Dragonfly topology, 200 Gb/s | Hewlett Packard Enterprise | - | Reported | pdc.kth.setop500.org |
The read
Sweden's HPE Cray EX at PDC, KTH Royal Institute of Technology, funded through NAISS and installed in stages between 2021 and 2023. PDC lists a CPU partition of 1,278 nodes (163,584 cores of AMD EPYC Zen 2) and a GPU partition of 62 nodes with four AMD Instinct MI250X each, on Slingshot with a Dragonfly topology (200 Gb/s after a March 2023 upgrade). The figures in this row are TOP500's for the GPU partition (9.90 PFlop/s Rmax, 10.94 PFlop/s Rpeak, 56,640 cores, 167.71 kW on the June 2026 list); the GPU partition first ranked 77th on the June 2023 list at 8.26 PFlop/s, and the separate CPU partition (5.06 PFlop/s Rmax) is listed on TOP500 as its own entry and is not summed in.
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
- 2026-10-01 Added Dardel (PDC Center for HPC)
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 and part | AMD EPYC 7742 (Rome), AMD Instinct MI250X, AMD EPYC 7A53 (Trento), HPE Slingshot-11 |
| pdc.kth.se | names system and part | AMD EPYC 7742 (Rome), AMD Instinct MI250X, AMD EPYC 7A53 (Trento) |
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.