NIPA(NHN Cloud)-CL-2 node architecture
NIPA(NHN Cloud)-CL-2 is operated by NHN Cloud in South Korea. 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
- #40
- HPL Rmax
- 68.42 PFlop/s
- Theoretical peak Rpeak
- 77.94 PFlop/s
- Ranked core count
- 320,280
Cluster · HPE ProLiant Compute XD685 · public profile
- Operator / site
- NHN Cloud
- Country
- South Korea
- Profile scope
- Ranked system / partition
- Data checked
- 14 September 2026
- Manufacturer
- HPE
- Ranked core count
- 320280
- Theoretical peak (PFlop/s)
- 77.94
- Operating system
- Ubuntu 24.04
HPE ProLiant Compute XD685 · public profile
Items marked ▧ come from the documented HPE ProLiant Compute XD685 platform specification.
- System / node model
- HPE ProLiant Compute XD685
- Compute nodes of this type
- 255
- CPU model
- AMD EPYC 9365 36C 3.4GHz
- CPU sockets per node
- 2
- Cores per socket
- 36
- Physical CPU cores
- 72
- Accelerator model
- NVIDIA B200
- Physical accelerators per node
- 8
- Host / unified memory
- 2,304 GB
- Interface speed
- 400 Gb/s (InfiniBand)
- Form factor
- 5U direct-liquid-cooled or 6U air-cooled chassis · 6 × 3000 W Titanium Plus (N+1)
Scope and uncertainty
Korea's national AI computing portal, run by NIPA, publishes the per-server configuration for this cluster: up to eight B200 accelerators, a 72-core CPU complement and 2,304 GB of memory per server, with up to 400 Gb/s InfiniBand between servers. NHN states the cluster comprises 255 nodes and 2,040 B200 GPUs. It publishes no per-node CPU, memory or adapter detail; accelerators per node follow the documented HPE ProLiant XD685 chassis. Node specifications follow the documented HPE ProLiant Compute XD685 platform (5U direct-liquid-cooled or 6U air-cooled chassis, two sockets, eight accelerators); operator-specific deviations are not published.
Documented fields: 15 / 29
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
HPE ProLiant Compute XD685 · public profile — 11 / 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 →
CLUSTERATLAS