Supercomputers/Systems/Deepthought2
USA · operational · hpc
Deepthought2
Also known as UMD Deepthought2
Operated by University of Maryland .
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 1581 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
- 163.67 TFlop/s
- Rpeak
- 298.21 TFlop/s
- Rmax ÷ Rpeak
- 54.9%
- Cores
- 10,160
- 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.
How it compares
Ranked against the other machines in this dataset that report the same figure: those that first came into service within two years of Deepthought2, and those in USA. Computed from our tables; a metric is ranked only when at least eight peers report it.
| Metric | Value | Against its era | Against its country |
|---|---|---|---|
| Rmax | 163.67 TFlop/s | #219 of 245 · bottom 12% | #237 of 332 · bottom 30% |
| Rmax ÷ Rpeak | 54.9 % | #226 of 243 · bottom 8% | #296 of 331 · bottom 12% |
Every metric is ranked largest-first, so a high power-draw rank means a big power draw, not a good one. Rmax ÷ Rpeak uses HPL-measured results only. Rankings move as machines are added; they describe this dataset, not the world.
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 | - Dell C8220/C8220X cluster | Dell Technologies | - | Reported | top500.orglabmanager.com |
| Accelerator | - NVIDIA K20 | NVIDIA | - | Reported | top500.org |
| CPU | - Intel Xeon E5-2680v2 10C 2.8GHz | Intel | - | Reported | top500.org |
| Interconnect | Mellanox InfiniBand FDR Interconnect Fat tree · 56 Gb/s per 4x port · About 1 microsecond Infiniband FDR | Mellanox NVDANASDAQ
rolls up to NVIDIA | - | Reported | top500.orglabmanager.com |
The read
Deepthought2 is the University of Maryland's Dell cluster, unveiled in May 2014 as about ten times faster than the original Deepthought of 2006 and housed in the Cyberinfrastructure Center, a 9,000 square foot facility in the Rivertech Building at M Square Research Park. The university's release gives about 300 TFlop/s, 1 PB of storage and an InfiniBand network. TOP500 lists the Dell C8220/C8220X cluster with Xeon E5-2680v2 10C 2.8 GHz, InfiniBand FDR and NVIDIA K20 GPUs at 10,160 cores, 163.67 TFlop/s Rmax and 298.21 TFlop/s Rpeak, 346th in June 2014 and 425th in November 2014.
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 Deepthought2 added from its source pages
Nearby in scale
The machines closest to Deepthought2 by Rmax, smaller to larger.
- Firestorm · Denmark, operational · 0.16 PFlop/s
- Grifo06 · Brazil, operational · 0.16 PFlop/s
- Diamond (ERDC) · USA, operational · 0.16 PFlop/s
- Taito (CSC) · Finland, operational · 0.16 PFlop/s
- Kyushu FX10 · Japan, operational · 0.17 PFlop/s
- Ivanoe · France, operational · 0.17 PFlop/s
Related systems
Same operator
- Zaratan · USA, operational, 2022
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.