ASRock Rack 1U2N2G-MILAN/2T GPU Server Barebone

The ASRock Rack 1U2N2G-MILAN/2T is a high-density dual-node 1U GPU server barebone engineered for data center AI inference and HPC deployments. The chassis houses two independent server nodes within a single 1U rack unit, each node built on the AMD EPYC Milan (3rd Gen) processor platform. Each node supports dual AMD EPYC 7003-series processors, delivering substantial per-node CPU core counts and memory bandwidth to feed attached GPU workloads efficiently. The dual 25GbE (2T) onboard networking per node enables high-throughput data ingestion pipelines essential for real-time inference serving.

Each node accommodates GPU accelerators via PCIe expansion, making the platform well-suited for deploying GPU-accelerated inference at rack scale with dramatically improved compute density compared to conventional 1U single-node designs. The barebone configuration ships without CPUs, memory, storage, or GPUs, giving data center architects full flexibility to configure each node independently according to workload requirements. Shared 1U chassis power and cooling infrastructure reduces per-node overhead while maintaining node-level operational independence.

Targeted at AI inference clusters, edge data centers, and colocation environments where rack space carries a premium, the 1U2N2G-MILAN/2T allows operators to double node density relative to traditional 1U server deployments. The Milan EPYC platform brings PCIe 4.0 connectivity, high memory channel counts per node, and proven RAS features expected in enterprise and HPC-grade infrastructure, making this barebone a strong foundation for GPU-accelerated workloads launched into production in Q4 2024.

Ideal for

• High-density AI inference serving in colocation and hyperscale data centers where rack unit space is constrained
• Dual-node GPU cluster deployments for natural language processing and computer vision inference at scale
• HPC workload consolidation requiring AMD EPYC Milan CPU performance paired with GPU acceleration in minimal rack footprint
• Edge data center AI inference nodes where power and space budgets demand maximum compute density per 1U
• Multi-tenant GPU inference infrastructure where node-level isolation within a shared 1U chassis is operationally advantageous
• Research and enterprise lab environments building cost-efficient GPU inference clusters with flexible barebone configurations

Technical specifications

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