Supercomputers/Systems/SWARM (Wuhan University)
China · operational · hpc
SWARM (Wuhan University)
Also known as SWARM
Operated by Wuhan University .
Researched with AI assistance, checked by scripts. These rows were researched by AI agents reading public operator, laboratory and vendor sources, appended by scripts and checked by automated validation; no subject-matter expert has reviewed them (how this site is made). A mechanical second-reader pass has since fetched every cited source: 1350 of 1540 systems have a readable citation that names them, and 64 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
- 1.40 PFlop/s
- Rpeak
- 3.07 PFlop/s
- Rmax ÷ Rpeak
- 45.6%
- Cores
- 32,640
- 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 | - Supermicro/Dell EMC C4130/1028GQ GPU servers | Supermicro | - | Verified | hpc.whu.edu.cntop500.org |
| Integrator | - Supermicro/Dell EMC C4130/1028GQ GPU servers (TOP500 manufacturer SuperMicro/DellEMC) | Dell Technologies | - | Verified | hpc.whu.edu.cntop500.org |
| Accelerator | NVIDIA Tesla V100 Accelerators CUDA (Volta GV100) · 5,120C · 1.53 GHz · 12 nm · 300 W Four NVIDIA Tesla V100 16GB per GPU node across 125 nodes (500 GPUs); TOP500 lists Tesla V100 SXM2 | NVIDIA | 500 reported | Reported | hpc.whu.edu.cntop500.org |
| CPU | - Two Intel Xeon E5-2640 v4 (10 cores, 2.4 GHz) per GPU node across 125 nodes | Intel | 250 reported | Reported | hpc.whu.edu.cntop500.org |
| Interconnect | Omni-Path Architecture Interconnect Fat tree · 100 Gb/s per port · About 1 microsecond 100G Omni-Path (page: 100G OPA; TOP500: Intel Omni-Path 100G) | Intel | - | Verified | hpc.whu.edu.cntop500.org |
| Operating system | CentOS Operating systems Linux distribution, RHEL-compatible rebuild · Open source, community-maintained CentOS | The CentOS Project | - | Reported | top500.org |
The read
GPU cluster of the Wuhan University supercomputing centre. TOP500 lists SWARM as Supermicro/Dell EMC C4130/1028GQ servers with Xeon E5-2640v4 10-core 2.4 GHz, Intel Omni-Path 100G and NVIDIA Tesla V100 SXM2 under CentOS, installed 2019, with 32,640 cores, Rmax 1.40 and Rpeak 3.07 PFlop/s and no power figure; it ranked 297th on its first list in June 2019 and 479th on its last, November 2020. The centre's resources page (in Chinese) lists a gpu partition of 125 Supermicro/Dell GPU servers, each with two Xeon E5-2640 v4 (20 cores) and four Tesla V100 16 GB (500 GPUs) on 100G Omni-Path, alongside Omni-Path CPU partitions (30 Lenovo SD530 and 28 Supermicro H240 nodes with Xeon Gold 6248 and 6248R, and 168 Xeon Phi 7250 KNL nodes) and several InfiniBand CPU partitions. The page does not use the name SWARM, gives no performance figure and is undated, so the match between the TOP500 entry and the 125-node partition rests on identical CPU, GPU and fabric; the tier is reported. The operator is the university's supercomputing centre; no street address or in-service date is stated, so the site is left blank.
Timeline
Public list appearances
Pointers only: rank and edition, linking to the canonical entry. We do not reproduce list tables. See sourcing policy.
Change history
- 2026-10-18 Added SWARM (Wuhan University)
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.