Skip to content
Back to product
N
NexusCompute

Hardware Datasheet · GPU Server

Nexus Compute

Nexus Compute

Nexus Compute HGX B200 8-GPU EPYC Inference Server

EPYC-driven Blackwell density for high-throughput, low-latency model serving.

Overview

This server pairs the NVIDIA HGX B200 8-GPU platform with dual AMD EPYC processors and abundant PCIe lanes, tuned for high-concurrency inference where memory bandwidth and tokens-per-second matter most. Nexus Compute specifies, configures, tests, and warranty-backs each system through authorized channels, optimizing GPU partitioning and networking around your latency and volume targets.

Specifications

GPUNVIDIA HGX B200 8-GPU (180GB HBM3e each, 1.4TB total)
GPU Interconnect5th-gen NVLink + NVSwitch, 1.8TB/s per GPU
CPUDual AMD EPYC (9004/9005 series)
System MemoryUp to 3TB DDR5 ECC
StorageHot-swap NVMe array + M.2 boot
Networking8x ConnectX-7 up to 400Gb/s (1:1 GPU:NIC)
GPU PartitioningMulti-instance partitioning for multi-tenant serving
Form Factor8U–10U rackmount (air-cooled)
WarrantyEnterprise warranty with support options

Typical Use Cases

  • ·Production LLM and generative model serving
  • ·High-concurrency, latency-sensitive inference
  • ·Multi-tenant GPU serving with partitioning
  • ·Retrieval-augmented generation at scale

Industries

SaaS & SoftwareAI & Machine LearningFinancial ServicesTelecom

Warranty & Support

Supplied through authorized channels with full manufacturer warranty. On-site, next-business-day support options available. Every system is configured, tested, and documented before delivery, with asset and warranty records provided for enterprise audit requirements.

Request a tailored quote

Configurations are tailored per engagement — contact us for pricing and lead times.

sales@nexus-compute.com

+1 737 276 1016

nexus-compute.com

Specifications are indicative and configured to each engagement. All product names, logos, and trademarks are the property of their respective owners. Nexus Compute is an independent enterprise hardware supplier.