Parts/CPUs/Intel Xeon E5-2680 v4
CPUs · Broadwell-EP · 2016
Intel Xeon E5-2680 v4
14-core Xeon host in the SGI ICE XA nodes of TSUBAME3.0.
Why this part
Broadwell-EP, the last of the E5 naming scheme before Intel moved to Platinum and Gold. Pleiades at NASA Ames is the useful case here: it has been continuously rebuilt since 2008 across at least five Xeon generations without ever being decommissioned, which makes it invisible to any analysis that treats a supercomputer as a single dated object. Systems that are upgraded rather than replaced are a real gap in every public dataset, including this one.
Systems using this part
- Systems
- 10
- Rmax underneath
- 25.08 PFlop/s
- Units recorded
- 2,392
2 of 10 edges disclose a quantity
| System | Role | Country | Units | System Rmax | Supplied by | Rank |
|---|---|---|---|---|---|---|
| Electra 2018 | CPU | USA | 1,152 | 5.44 PFlop/s | Intel | #46 → |
| TSUBAME3.0 2017-08 | CPU | Japan | - | 8.13 PFlop/s | Intel | #59 → |
| Sator 2017 | CPU | France | 1,240 | 579.23 TFlop/s | Intel | #343 → |
| Unayzah (2017) 2017 | CPU | Saudi Arabia | - | 484.50 TFlop/s | Intel | #438 → |
| Biowulf 2016 | CPU | USA | - | 1.97 PFlop/s | Intel | #217 → |
| BrENIAC 2016 | CPU | Belgium | - | 548.00 TFlop/s | Intel | #197 → |
| Owens 2016 | CPU | USA | - | 705.97 TFlop/s | Intel | #273 → |
| BinAC 2016 | CPU | Germany | - | 334.80 TFlop/s | Intel | #414 → |
| Eole 2016 | CPU | France | - | 942.83 TFlop/s | Intel | #129 → |
| Pleiades 2008-11 | CPU | USA | - | 5.95 PFlop/s | Intel | #3 → |
“Units recorded” sums only the edges where a public source states a quantity. It is a floor, not a total, and should never be read as installed base.