NEC Express5800 R120h-1M GPU-Ready Rack Server

The NEC Express5800 R120h-1M (N8100-2840Y) is a 1U rack server engineered to deliver GPU-accelerated compute within a compact, data-center-ready form factor. Built on the Intel Xeon Scalable processor platform, the R120h-1M supports dual-socket CPU configurations and incorporates PCIe expansion designed to accommodate GPU accelerators, making it well-suited for enterprise AI inference, edge analytics, and high-density virtualization workloads where rack space efficiency is critical.

The server integrates NEC's mature Express5800 reliability engineering, including redundant power supply support, hot-swap storage bays, and out-of-band management via ESMPRO — NEC's enterprise systems management suite — enabling administrators to monitor, configure, and update the platform remotely without service interruption. Its balanced memory subsystem, supporting DDR4 RDIMM and LRDIMM configurations across multiple DIMM slots per socket, provides the memory bandwidth required to feed GPU accelerators during continuous inference operations.

Targeted at enterprise data centers modernizing their infrastructure for AI and machine learning inference pipelines, the R120h-1M fits within existing 1U rack allocations while adding GPU-ready expansion not typically found in single-unit servers of this density. NEC positions this platform within its 2025 refresh of the Express5800 R120h line, addressing demand from verticals such as financial services, healthcare, and manufacturing where low-latency inference at the edge of the data center is becoming a standard architectural requirement.

Ideal for

• AI and machine learning inference serving at data-center or near-edge tier with GPU acceleration in a 1U footprint
• High-density virtualization hosting requiring a balance of per-socket CPU core count, memory capacity, and accelerator support
• Real-time analytics and event-stream processing for financial services or manufacturing operations intelligence
• Healthcare imaging AI inference, enabling rapid processing of diagnostic image data within on-premises regulated environments
• HPC departmental workloads requiring GPU offload within space-constrained co-location or enterprise data halls
• Enterprise Kubernetes or container-based microservices platforms requiring GPU node integration for inference sidecar services

Technical specifications

Available from Omnixon Global. Submit an RFQ and our team will confirm configuration and availability for your order.