Hitachi Advanced Server DS225 (AMD EPYC Gen4)

The Hitachi Advanced Server DS225 is a purpose-built 2U form-factor system featuring dual AMD EPYC 9004-series processors (Bergamo architecture), delivering up to 384 cores and 768 threads per node. The platform integrates 12 DIMM slots per socket supporting up to 24 DDR5-5600 memory modules for maximum aggregate bandwidth and capacity, enabling memory-intensive HPC simulations, large language model inference, and analytics workloads. The DS225 incorporates PCIe 5.0 connectivity for accelerator interoperability, redundant 2000W power supplies, and comprehensive thermal management designed for continuous high-utilization environments.

Architecturally, the server leverages AMD EPYC 9004's 5nm process technology, 12-channel memory controllers, and advanced security features including encrypted virtualization and confidential computing. The platform supports hot-swap drive bays compatible with 2.5-inch SSD/HDD configurations, enabling flexible storage tiering for data-intensive applications. Native support for AMD SEV-SNP and TSME provides cryptographic isolation for sensitive workloads across shared infrastructure.

The DS225 addresses enterprise demand for consolidated compute density in cloud service provider environments, research institutions requiring exascale-class simulation capabilities, and organizations deploying distributed AI inference clusters. Launched in Q2 2024, the system represents Hitachi's commitment to next-generation processor architecture adoption and optimization for heterogeneous compute-accelerated environments.

Ideal for

• Large-scale HPC simulations (finite element analysis, computational fluid dynamics, molecular dynamics) benefiting from 384 cores and unified memory bandwidth
• Distributed AI model inference serving multiple concurrent requests with low-latency response requirements across containerized workloads
• Cloud service provider multi-tenant virtualization environments requiring secure compute isolation and high core-density consolidation
• Real-time analytics and data warehousing with TeraScale in-memory datasets leveraging DDR5 capacity and bandwidth
• Machine learning model training acceleration using PCIe 5.0-attached GPUs (H100, H200, MI300) for rapid iteration cycles
• Scientific research and genomics analysis requiring parallel processing of massive datasets with deterministic performance characteristics

Technical specifications

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