Supercomputers/Systems/Raad2
Qatar · operational · hpc
Raad2
Also known as RAAD2, Raad2 Cray XC40
Operated by Hamad Bin Khalifa University .
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: 1263 of 1345 systems have a readable citation that names them, and 56 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
- 120.00 TFlop/s
- Rpeak
- -
- Rmax ÷ Rpeak
- -
- Cores
- 4,128
- 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 | - Cray XC40 supplied by Cray (now HPE) | Cray HPENYSE
rolls up to Hewlett Packard Enterprise | - | Reported | rccg.hbku.edu.qarccg.hbku.edu.qa |
| Accelerator | NVIDIA Tesla V100 Accelerators CUDA (Volta GV100) · 5,120C · 1.53 GHz · 12 nm · 300 W NVIDIA V100 GPUs, eight on four external nodes (Raad2-GFX) | NVIDIA | 8 reported | Reported | rccg.hbku.edu.qa |
| CPU | - Intel Xeon Haswell, 24 cores per node, 4,128 cores | Intel | - | Reported | rccg.hbku.edu.qarccg.hbku.edu.qa |
| Interconnect | Cray Aries Interconnect Dragonfly · About 10 GB/s per direction · Around 1 microsecond Cray Aries interconnect in a Dragonfly topology | Cray HPENYSE
rolls up to Hewlett Packard Enterprise | - | Reported | rccg.hbku.edu.qa |
| Storage | - DDN ExaScaler 7990X Lustre file system, 800 TB | DDN | - | Reported | rccg.hbku.edu.qarccg.hbku.edu.qa |
The read
Qatar's flagship research supercomputer, a Cray XC40 acquired from Cray in 2016 and jointly bought by the Qatar Computing Research Institute and Texas A&M University at Qatar, now run by the Research Computing Center Group of Hamad Bin Khalifa University. The HBKU systems page lists 4,128 Intel Xeon Haswell cores at 24 cores and 128 GB per node on the Aries interconnect in a Dragonfly topology, 120+ TFLOPS of Linpack, and a DDN ExaScaler 7990X Lustre store of 800 TB; the Raad2-GFX extension adds eight NVIDIA V100 GPUs on four external nodes joined over FDR InfiniBand. The same page says 192 nodes while the history page says 172 compute nodes, and 172 nodes of 24 cores matches the 4,128 cores, so the node count is not recorded. The history page gives 2017 to the present, SUSE Linux with Slurm and 16 GB/s peak read bandwidth. It follows Saqr (2005 to 2008, Apple G5 Xserve cluster), Suqoor (2008 to 2013, 512-core SGI Altix) and the first Raad (2012 to 2017, about 2,200 Sandy Bridge cores and 12 K20 GPUs). The systems page also lists a Cray XC50 named Hazeem acquired in 2018 and moved to the same group in the second quarter of 2025; it is not recorded here for lack of specifications. Both fetched pages come from the same organisation and the system is not on the TOP500, so the tier is reported and the status date is the day the pages were read.
Timeline
- 2017 First operational source
Change history
- 2026-10-09 Added Raad2
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 |
|---|---|---|
| rccg.hbku.edu.qa | names system, part and count | Cray Aries, NVIDIA Tesla V100 · counts: 8 |
| rccg.hbku.edu.qa | names system, part and count | Cray Aries · counts: 8 |
Checked 2026-10-13 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.