CLUSTERATLAS
A field guide to supercomputing

Derecho

NSF NCAR · Wyoming Supercomputing Center

CPU Accelerator Interconnect
i
Reading this map

Derecho node architecture

Derecho is operated by NSF NCAR · Wyoming Supercomputing Center in United States. This profile covers Documented compute profile. Specifications are listed below; anything the operator has not published is itemised in the documented-fields list.

Cluster · CPU compute

Operator / site
NSF NCAR · Wyoming Supercomputing Center
Country
United States
Profile scope
Documented compute profile
Data checked
14 September 2026
Operating system
Cray variant of SUSE Linux Enterprise
Network topology
Dragonfly

CPU compute

Compute nodes of this type
2488
CPU model
AMD EPYC 7763 · 64 cores
CPU sockets per node
2
Cores per socket
64
Physical CPU cores
128
Host / unified memory
256 GB DDR4
Network interfaces per node
1
Interface model
HPE Slingshot 11
Interface speed
200 Gb/s per port per direction
Interface-to-accelerator topology
One Slingshot injection port per CPU-only node

A100 compute

Compute nodes of this type
82
CPU model
AMD EPYC 7763 · 64 cores
CPU sockets per node
1
Cores per socket
64
Physical CPU cores
64
Accelerator model
NVIDIA A100
Physical accelerators per node
4
Host / unified memory
512 GB DDR4
Accelerator memory
40 GB HBM2 per GPU
Accelerator ↔ accelerator link
NVLink · 600 GB/s GPU interconnect
On-node fabric
NVLink — 600 GB/s GPU interconnect
Network interfaces per node
4
Interface model
HPE Slingshot 11
Interface speed
200 Gb/s per port per direction
Interface-to-accelerator topology
Four Slingshot injection ports per GPU node

Scope and uncertainty

Production CPU and GPU compute nodes only; login and development nodes excluded.

Documented fields: 27 / 40

Every field this catalog checks. Fields listed above have a value; the remainder are shown here. A field marked absent is not publicly available in the sources we could verify, which does not mean the hardware is missing. Fields that cannot apply, such as benchmark results for an unranked system, are excluded from the total.

Cluster — 2 / 3

  • Manufacturer
  • Ranked core count
  • Theoretical peak (PFlop/s)
  • Measured power (kW)
  • Energy efficiency (GFlop/s per watt)
  • Operating system
  • Network topology

CPU compute — 10 / 15

  • System / node model
  • Compute nodes of this type
  • CPU model
  • CPU sockets per node
  • Cores per socket
  • Physical CPU cores
  • CPU note
  • Accelerator model
  • Physical accelerators per node
  • Host / unified memory
  • Accelerator memory
  • Accelerator ↔ accelerator link
  • On-node fabric
  • CPU ↔ accelerator link
  • Network interfaces per node
  • Interface model
  • Interface speed
  • Interface-to-accelerator topology
  • Aggregate scale-out per node
  • GPUDirect RDMA
  • Management / storage networking
  • Local storage
  • NVMe data disks
  • Max remote data disks
  • Uncached disk IOPS
  • Uncached disk throughput (MBps)
  • Form factor

A100 compute — 15 / 22

  • System / node model
  • Compute nodes of this type
  • CPU model
  • CPU sockets per node
  • Cores per socket
  • Physical CPU cores
  • CPU note
  • Accelerator model
  • Physical accelerators per node
  • Host / unified memory
  • Accelerator memory
  • Accelerator ↔ accelerator link
  • On-node fabric
  • CPU ↔ accelerator link
  • Network interfaces per node
  • Interface model
  • Interface speed
  • Interface-to-accelerator topology
  • Aggregate scale-out per node
  • GPUDirect RDMA
  • Management / storage networking
  • Local storage
  • NVMe data disks
  • Max remote data disks
  • Uncached disk IOPS
  • Uncached disk throughput (MBps)
  • Form factor

Sources

Browse all cluster architecture pages →

Data & sources