Lenovo Adds AMD EPYC Servers and AI Factory Services for HPC - Unite.AI

Lenovo is treating AI and HPC as a system-level infrastructure challenge. As continuous inference, agentic execution and scientific workloads grow, performance depends on compute and memory bandwidth as well as storage, data movement and cooling. Scott Patti, vice president of Lenovo's Infrastructure Solutions Group, said the goal is "getting more useful performance from every watt," while addressing power and cooling constraints and moving faster from deployment to operational AI.

Liquid-Cooled Servers

The dual-socket ThinkSystem SC755 is the Neptune liquid-cooled server in the update. Lenovo's internal testing reports that it can remove up to 100% of system heat through water and deliver up to 40% higher sustained performance than similar air-cooled systems. Those results vary with system configuration, workload, facility water temperature and data-center conditions.

Dynamic Partitioning lets customers convert two dual-socket systems into single sockets and back, making compute allocation more flexible across workloads. For compute density and memory bandwidth, Lenovo is adding the ThinkSystem SR755, which targets agentic AI infrastructure, electronic design automation, financial modeling, engineering and large-scale virtualization.

Storage and Interconnects

Two storage systems extend the focus on data throughput. The ThinkSystem DE6800 is intended for applications requiring reliable, high-throughput storage, while the DE6500 offers economical performance with simpler deployment and management. Both address latency-sensitive block-storage workloads.

Validation of the Cornelis CN5000 fabric also gives customers another high-performance interconnect intended to reduce data-movement bottlenecks in demanding AI and HPC environments.

AI Factory Services

The services portion addresses the work after hardware arrives. Hybrid AI Factory Services now cover GPU deployment, configuration and optimization. Multi-node support for Red Hat AI Enterprise and NVIDIA AI Enterprise with NVIDIA Run:ai provides two software paths for deploying, governing and scaling workloads, including agentic AI and physical AI.

At the larger end, Lenovo's expanded Gigafactory Services cover the concept and build phases through deployment and optimization for production. NVIDIA Infrastructure Services expertise is included for specialized deployment and cluster validation. Lenovo can also facilitate access to Digital Realty's colocation network and the Digital Realty Innovation Lab.

Customer Deployments

The University of Florida's fourth-generation HiPerGator cluster combines Lenovo, NVIDIA and AMD technology. Lenovo's case study lists 60,000 AMD EPYC server CPU cores and 1,552 NVIDIA GPUs and gives a one-year total of 33 million research requests. The TOP500 entry for HiPerGator AI lists 17.49 PFlop/s of measured Linpack Rmax performance and a theoretical peak of 19.00 PFlop/s.

DreamWorks Animation replaced a traditional air-cooled render farm with Lenovo ThinkSystem servers using Neptune liquid cooling. Lenovo's case study reports up to 100% effective heat removal, a 6-degree increase in the chilled-water set point and up to 10% lower annual data-center power costs, adding that the higher set point reduces energy use and creates headroom for additional liquid-cooled infrastructure.