Reindeer

Microsoft Azure

Node type

CPU Accelerator Interconnect

Reading this map

Reindeer node architecture

Reindeer is operated by Microsoft Azure in United States. This profile covers Ranked system / partition. Specifications are listed below; anything the operator has not published is itemised in the documented-fields list.

Published benchmark (TOP500 · June 2026)

HPL rank: #57
HPL Rmax: 45.59 PFlop/s
Theoretical peak Rpeak: 56.42 PFlop/s
Ranked core count: 138,240

Cluster · ND H200 v5

Operator / site: Microsoft Azure
Country: United States
Profile scope: Ranked system / partition
Data checked: 14 September 2026
Manufacturer: Microsoft
Ranked core count: 138240
Theoretical peak (PFlop/s): 56.42
Operating system: Ubuntu 22.04

ND H200 v5

System / node model: Microsoft ND H200 v5 (Azure Standard_ND96isr_H200_v5)
CPU model: Intel Xeon Platinum 8480C (Sapphire Rapids)
CPU sockets per node: 96
Physical CPU cores: 96
CPU note: Microsoft lists 96 vCPUs for the ND H200 v5 size; the physical host socket layout behind that allocation is not published.
Accelerator model: NVIDIA H200
Physical accelerators per node: 8
Host / unified memory: 1,850 GiB (VM specification)
Accelerator memory: 141 GB per GPU
Accelerator ↔ accelerator link: NVLink 4 · all GPU pairs
On-node fabric: NVLink 4 within the VM
Network interfaces per node: 8
Interface model: NVIDIA Quantum-2 CX7 InfiniBand
Interface speed: 400 Gb/s per adapter
Interface-to-accelerator topology: One dedicated, topology-agnostic adapter per GPU
Aggregate scale-out per node: 3.2 Tb/s per VM
GPUDirect RDMA: Yes
Local storage: 1,024 GiB temporary disk (VM specification)
NVMe data disks: Up to 8 × 28 TiB NVMe data disks (VM specification)
Max remote data disks: 16
Uncached disk IOPS: 40800
Uncached disk throughput (MBps): 612

Scope and uncertainty

The node model uses public ND H200 v5 VM specifications. TOP500 ranks this system #57 (June 2026) with a Rmax of 45.59 PFlop/s. Exact benchmark VM count, the physical host socket allocation behind the 96 reported vCPUs, and the scale-set topology are not published.

Documented fields: 26 / 34

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 — 4 / 7

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

ND H200 v5 — 22 / 27

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 →