Supercomputers/Systems/HKUST SuperPOD
China · operational · mixed
HKUST SuperPOD
Operated by HKUST .
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 |
|---|---|---|---|---|---|
| Integrator | - NVIDIA DGX SuperPOD H800 systems, 55 GPU nodes | NVIDIA | - | Reported | itso.hkust.edu.hkitso.hkust.edu.hk |
| Accelerator | - NVIDIA H800, eight per DGX H800 node across 55 nodes | NVIDIA | 440 estimated | Reported | itso.hkust.edu.hk |
| CPU | - Dual 56-core Intel Xeon processors per DGX node, 6,160 CPU cores in total | Intel | - | Reported | itso.hkust.edu.hk |
| Interconnect | NVIDIA InfiniBand NDR (Quantum-2) Interconnect Fat tree or dragonfly+ · 400 Gb/s per port · Under 0.6 microseconds NVIDIA NDR (400 Gbps) InfiniBand, 400 Gb/s connections per node | NVIDIA | - | Reported | itso.hkust.edu.hkitso.hkust.edu.hk |
| Storage | - 500 TB DDN AI400X2 storage (tier 1) | DDN | - | Reported | itso.hkust.edu.hk |
| Storage | - 5.7 PB Dell PowerScale storage (tier 2) | Dell Technologies | - | Reported | itso.hkust.edu.hk |
The read
The Hong Kong University of Science and Technology's central AI-for-Science facility, built on NVIDIA DGX SuperPOD H800 systems: 55 GPU nodes each with eight H800 GPUs (440 in all), dual 56-core Intel Xeon CPUs, 2 TB of RAM and 400 Gb/s InfiniBand per node, with 6,160 CPU cores in total, a 500 TB DDN AI400X2 flash tier and a 5.7 PB Dell PowerScale tier, run under SLURM at 7x24 service. HKUST quotes 448 PFlop/s of tensor performance and 30 PFlop/s standard floating point without naming the precision, so no FLOPS figure is recorded. The pages give no launch date or machine-room location. Hong Kong is recorded under China to match the dataset's country list. Both cited pages are from HKUST's IT office, so the tier is reported.
Timeline
No dated events recorded.
Change history
- 2026-10-02 HKUST SuperPOD added, from the HKUST IT office
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 |
|---|---|---|
| itso.hkust.edu.hk | names system and part | NVIDIA InfiniBand NDR (Quantum-2) |
| itso.hkust.edu.hk | names system and part | NVIDIA InfiniBand NDR (Quantum-2) |
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
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.