Supercomputers/Systems/Taiwania 3
Taiwan · operational · hpc
Taiwania 3
Operated by National Center for High-Performance Computing at NCHC Hsinchu .
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
- 2.30 PFlop/s
- Rpeak
- -
- Rmax ÷ Rpeak
- -
- Cores
- 50,400
- Power
- 563.85 kW
- Per watt
- 4.1 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 | - QCT QuantaGrid T42D-2U | Quanta Cloud Technology | - | Verified | blog.qct.iotop500.org |
| CPU | Intel Xeon Platinum 8280 CPUs x86-64 · 28C · 2.7 GHz · 14 nm · 205 W two Xeon Platinum 8280 per node, 900 nodes | Intel | 1,800 estimated | Verified | nchc.org.twtop500.org |
| Interconnect | NVIDIA InfiniBand HDR Interconnect Fat tree or dragonfly+ · 200 Gb/s per port · About 0.6 microseconds InfiniBand HDR100 100 Gbps | NVIDIA | - | Verified | nchc.org.twtop500.org |
| Operating system | CentOS Operating systems Linux distribution, RHEL-compatible rebuild · Open source, community-maintained CentOS | The CentOS Project | - | Reported | top500.org |
The read
NCHC's CPU-only cluster of 900 nodes, each with two 28-core Xeon Platinum 8280 CPUs, on HDR100 InfiniBand, built with QCT QuantaGrid servers. It first appeared on the November 2020 TOP500 at 181st and had slid to 456th by June 2024, its last listing. NCHC and QCT give a peak of 2.7 PFlop/s where TOP500 says 4.35, so no peak is recorded and the conflict is logged. TOP500 gives the clock as 2.7 GHz and NCHC as 2.4 GHz. No official start date was found in a page we fetched, so it is blank. NCHC's current supercomputer page still lists the system, but it is undated, so the operational status is a reading of that page. TOP500 is the only source for Rmax, Rpeak and core count, so the tier is reported.
Timeline
- 2020-11 List debut at #181 (November 2020) source
Public list appearances
Pointers only: rank and edition, linking to the canonical entry. We do not reproduce list tables. See sourcing policy.
Site & facility
Where sources disagree
We record conflicts instead of picking a winner quietly. All open questions.
- Peak performance: 4.35 PFlop/s (TOP500 Rpeak) against 2.7 PFlop/s (NCHC). QCT also gives 2.7. No peak is recorded.
Change history
- 2026-09-20 Added with a rank pointer; its peak is disputed between TOP500 and NCHC
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 |
|---|---|---|
| nchc.org.tw | names system and part | Intel Xeon Platinum 8280, NVIDIA InfiniBand HDR, CentOS |
| top500.org | names system and part | Intel Xeon Platinum 8280, NVIDIA InfiniBand HDR, CentOS |
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.