The new systems target AI inference, virtualization, analytics and high-performance computing, combining next-generation AMD EPYC processors with automated security and simplified infrastructure management.
Hewlett Packard Enterprise (HPE) has introduced the first four HPE ProLiant Gen13 servers powered by 6th Generation AMD EPYC processors, expanding its enterprise compute portfolio for AI-driven workloads and data-intensive applications.
The systems are designed to support agentic AI, inference, analytics, virtualization, simulation and modelling while improving performance, security and operational efficiency across distributed IT environments.
The launch comes as enterprises look to scale AI beyond experimentation. HPE cited McKinsey’s 2026 AI Trust Maturity Survey, which found that nearly two-thirds of respondents identified security and risk concerns as the leading barrier to scaling agentic AI.
The new portfolio integrates HPE Integrated Lights-Out 8 (iLO 8) security capabilities with HPE Compute Ops Management to strengthen infrastructure protection and simplify deployment, monitoring and lifecycle management.
The HPE ProLiant DL585a Gen13 is designed for demanding AI workloads, supporting up to eight double-wide PCIe GPUs, two AMD EPYC processors and PCIe Gen6 connectivity. It targets AI inference, retrieval-augmented generation fine-tuning, analytics and virtualization.
“Organizations need to turn their investments in agentic AI and inference into business value without adding operational complexity or risk.” — Krista Satterthwaite, Senior Vice President and General Manager, Compute, HPE
The compact, single-socket HPE ProLiant DL525 Gen13 supports up to 256 CPU cores and 16-channel memory, making it suitable for high-throughput workloads such as fraud detection, financial modelling, electronic design automation and inference.
For compute-intensive applications, the HPE ProLiant XD245 Gen13 uses a four-node, direct-liquid-cooled design, while the XD285 Gen13 adopts a two-node, air-cooled architecture. Both target high-performance computing, scientific research, engineering simulations and climate modelling.
Security enhancements include automatic self-encrypting drive capabilities, multi-person authorization for sensitive actions, expanded post-quantum cryptography readiness and component verification built on HPE’s Silicon Root of Trust.
HPE says customers using Compute Ops Management spent 50% less time managing servers and generated 209% ROI, based on a commissioned Forrester Consulting study.
The DL525 Gen13 is expected later in 2026, the DL585a in early 2027, and the XD245 and XD285 in 2027. The rollout extends HPE’s strategy of combining performance, security and operational simplicity as enterprises modernize infrastructure for AI and traditional workloads.