Lenovo on Tuesday announced two ThinkSystem servers built on AMD's 6th Generation EPYC 9006 processors, two storage arrays and expanded AI services ahead of the SC26 supercomputing conference Lenovo StoryHub1.

The new lineup targets organizations running complex AI and HPC workloads such as continuous inference, agentic execution and scientific discovery across on-premises, hybrid and colocation environments.

Lenovo's Hybrid AI strategy spans Personal AI and Enterprise AI, positioning the company as a full-stack provider across devices, infrastructure and services. The announcement builds on the Lenovo AI Cloud Gigafactory program with NVIDIA, introduced in January 2026, which emphasizes time to first token as a deployment benchmark Lenovo StoryHub2.

Servers and storage

The lineup includes two ThinkSystem servers and two storage arrays:

  • ThinkSystem SC755: A dual-socket Neptune liquid-cooled server designed to remove up to 100% of system heat via water and deliver up to 40% higher sustained performance compared to similar air-cooled systems, based on Lenovo internal testing Lenovo StoryHub1. Lenovo Dynamic Partitioning enables flexible allocation of compute resources across workloads by converting two dual-socket systems into single sockets and back.

  • ThinkSystem SR755: Engineered for high-bandwidth workloads such as large-scale virtualization, engineering, financial modeling, electronic design automation and agentic AI infrastructure. The system targets higher compute density and memory bandwidth with lower power consumption than dual-socket configurations, based on Lenovo internal analysis Lenovo StoryHub1.

  • ThinkSystem DE6800 and DE6500: Storage arrays optimized for high-throughput, latency-sensitive block workloads. The DE6800 is positioned for high-throughput applications demanding reliable storage, while the DE6500 offers economical high performance with simplified deployment and management Lenovo StoryHub1.

Lenovo also highlights the Cornelis CN5000 fabric validation as a high-performance interconnect option for demanding AI and HPC environments, aimed at reducing bottlenecks across data-intensive workloads Lenovo StoryHub1.

Services and colocation

Lenovo is expanding its Hybrid AI Factory Services to help customers move from infrastructure deployment to operational AI capacity faster. New capabilities include GPU deployment, configuration and optimization, multi-node support across Red Hat AI Enterprise and NVIDIA AI Enterprise with NVIDIA Run:ai, and support for agentic AI and physical AI workloads Lenovo StoryHub1.

The company is also introducing end-to-end gigafactory services, including NVIDIA Infrastructure Services expertise for specialized deployment and cluster validation. For organizations needing additional data center capacity, Lenovo now facilitates access to Digital Realty’s global colocation facilities, with Lenovo providing a single point of contact and scalable infrastructure through Lenovo TruScale Infrastructure as a Service Lenovo StoryHub1.

Lenovo said existing customer deployments illustrate performance claims. At the University of Florida, the fourth generation of HiPerGator, built on Lenovo, NVIDIA and AMD technology, delivers 17.49 PFLOPS and processed 33 million research requests in the past year. DreamWorks Animation leverages Lenovo Neptune liquid-cooled servers to deliver a 20% performance improvement and faster rendering for creative iteration, according to Lenovo Lenovo StoryHub1.

What Lenovo has not disclosed

The announcement does not include independent test results or pricing. Lenovo states that heat removal and performance results vary by system configuration, workload, facility water temperature and data center conditions, and that actual results may vary Lenovo StoryHub1.

The press release is Lenovo's own announcement and includes no independent verification of its performance claims.

The announcement does not state availability dates or pricing.