Supercomputers/Systems/Jaguar
USA · upgraded · hpc
Jaguar
Also known as Cray XT5-HE
Operated by Oak Ridge National Laboratory at Oak Ridge Leadership Computing Facility .
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
- 1.76 PFlop/s
- Rpeak
- 2.33 PFlop/s
- Rmax ÷ Rpeak
- 75.5%
- Cores
- 224,162
- Power
- 6.95 MW
- Per watt
- 0.3 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 Jaguar, 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 | 1.76 PFlop/s | #4 of 136 · top 2% | #154 of 332 · top 46% |
| Energy efficiency | 0.3 GF/W | #45 of 84 · bottom 48% | #144 of 162 · bottom 12% |
| Rmax ÷ Rpeak | 75.5 % | #98 of 136 · bottom 29% | #158 of 331 · top 48% |
| Power draw | 6,950 kW | #2 of 86 · top 1% | #9 of 171 · top 5% |
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 | - Cray XT5-HE | Cray HPENYSE
rolls up to Hewlett Packard Enterprise | - | Reported | ornl.gov |
| CPU | AMD Opteron 2435 (Istanbul) CPUs x86-64 · 6C · 2.6 GHz · 45 nm · 75 W AMD Opteron six-core 2.6 GHz | AMD | - | Reported | ornl.gov |
| Interconnect | Cray SeaStar2+ Interconnect 3D torus · 9.6 GB/s per direction Cray SeaStar2+ | Cray HPENYSE
rolls up to Hewlett Packard Enterprise | - | Verified | ornl.govtop500.org |
| Storage | Lustre Storage Parallel filesystem, separate metadata and object servers Spider, Lustre | Cray HPENYSE
rolls up to Hewlett Packard Enterprise | - | Reported | ornl.govolcf.ornl.gov |
| Cooling | Liquid-assisted air (in-cabinet coil) Cooling In-cabinet air, water-cooled coil · Roughly 20-45 kW per cabinet Cray XT5 cabinet: liquid-assisted air | Cray HPENYSE
rolls up to Hewlett Packard Enterprise | - | Reported | ornl.govolcf.ornl.gov |
| Operating system | Cray Linux Environment Operating systems Linux, Cray-customised · Proprietary layer over open-source Linux Cray Linux Environment | Cray HPENYSE
rolls up to Hewlett Packard Enterprise | - | Reported | ornl.gov |
Engineering, funding and access
What Jaguar runs on, who paid for it and how people get time on it. Each line carries who said it and when.
Hardware
| Item | Detail | As of | Basis | Source |
|---|---|---|---|---|
| Total memory | 362 TBSystem memory across the XT5 and XT4 components at the November 2009 announcement. | 2009-11 | Operator stated | ornl.gov |
| Storage capacity | 10 PBSpider shared file system. | 2009-11 | Operator stated | ornl.gov |
| Interconnect and topology | Cray SeaStar network interface and router; an InfiniBand network connects the XT5 and XT4 components | 2009-11 | Operator stated | ornl.gov |
Software
| Item | Detail | As of | Basis | Source |
|---|---|---|---|---|
| Parallel filesystem | Lustre-based Spider shared file system, 10 PB, 284 GB/s I/O bandwidth | 2009-11 | Operator stated | ornl.gov |
Funding
| Item | Detail | As of | Basis | Source |
|---|---|---|---|---|
| Funded by | US Department of Energy Office of Science | 2009-11 | Operator stated | ornl.gov |
| Funding programme | American Recovery and Reinvestment Act (2009 six-core upgrade) | 2009-11 | Operator stated | ornl.gov |
| Stated cost | 19.9 USD mCost of the 2009 XT5 upgrade only, funded under the Recovery Act; not the whole machine. | 2009-11 | Operator stated | ornl.gov |
Results
| Item | Detail | As of | Basis | Source |
|---|---|---|---|---|
| Notable result | GE ran its largest-ever CFD calculation on Jaguar, its first unsteady turbomachinery study at that scale (2012)Accessed through ORNL's HPC Industrial Partnerships Program; GE later bought its own Cray system. | 2012-06 | Operator stated | olcf.ornl.gov |
| Notable result | First place on the November 2009 TOP500 with 1.759 PFlop/s HPL on the XT5 component (2009)The XT4 component was 16th on the same list at 0.205 PFlop/s. | 2009-11 | Operator stated | ornl.gov |
The read
CPU-only, Opteron-based, on Cray's SeaStar2+ torus. Notable mainly for what happened next: rather than being decommissioned it was rebuilt in place into Titan, making the Jaguar-to-Titan transition the first documented case of an accelerator generation displacing a CPU-only design at the top of the installed base. Cray was an independent public company at the time; HPE acquired it in 2019.
Timeline
- 2005 Announced source
- 2005 Jaguar entered service with a peak of 26 teraflop/s source
- 2009-11 List debut at #1 (November 2009) source
- 2009-11 First operational source
- 2009-11 Announced at SC09 as the world's fastest supercomputer after the Cray XT5 reached 1.759 petaflop/s on HPL source
- 2009-11 XT5 moved from quad-core to six-core AMD Istanbul processors (224,256 cores) in a USD 19.9 million Recovery Act upgrade source
- 2012-10 Rebuilt in place as Titan: Opteron 6274 nodes with one NVIDIA K20X each 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
Oak Ridge Leadership Computing Facility
- Location
- Oak Ridge, TN, USA
- Utility
- Tennessee Valley Authority
- Cooling
- Direct liquid cooling (warm water)
- Commissioned
- 1993
Where sources disagree
We record conflicts instead of picking a winner quietly. All open questions.
- Core count: 224,162 as recorded (TOP500) against 224,256 XT5 cores, 255,584 including the XT4 partition. ORNL counts partitions the TOP500 entry does not.
Nearby in scale
The machines closest to Jaguar by Rmax, smaller to larger.
- CARA · Germany, operational · 1.75 PFlop/s
- Quriosity (2017) · Germany, decommissioned · 1.75 PFlop/s
- Tryton Plus · Poland, operational · 1.73 PFlop/s
- G42 POD3 · United Arab Emirates, operational · 1.76 PFlop/s
- RE1A (Flatiron) · USA, operational · 1.77 PFlop/s
- Fermi · Italy, decommissioned · 1.79 PFlop/s
Related systems
Same site
- Titan · USA, decommissioned, 2012
- Phoenix (ORNL) · USA, decommissioned, 2003
- Cheetah (ORNL) · USA, decommissioned, 2002
- Summit · USA, decommissioned, 2018
- Andes (OLCF) · USA, operational, 2020
- Frontier TDS · USA, operational, 2021
Same operator
- Frontier · USA, upgraded, 2022
- Intel Paragon XP/S 150 · USA, decommissioned, 1995
- Discovery (ORNL) · USA, planned
- Lux (ORNL) · USA, planned
- ORNL Paragon XP/S 35 · USA, operational, 1992
- Crusher · USA, decommissioned
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 |
|---|---|---|
| olcf.ornl.gov | names system and part | Cray Linux Environment |
| ornl.gov | could not be fetched | - |
Checked 2026-10-18 by pnpm verify. Re-run it and the table changes with the web.
Further reading
Power and cooling
Why air ran out at 20 kW a rack, how warm-water cooling removes the chillers, and what PUE conceals.
Inside a node
Latency-optimised versus throughput-optimised processors, NUMA, and the end of the host-to-device copy.
Operating a supercomputer
Acceptance testing, why "installed" and "in production" are six months apart, and the machine lifecycle.