Harnesses

Lenovo launches new generation of high-performance infrastructure to accelerate AI and HPC convergence

November 18, 2026 · 3 min read

two black flat screen computer monitors
Fotis Fotopoulos / Unsplash

Lenovo has announced the launch of a new generation of products and services designed to support the rapid convergence of artificial intelligence and high-performance computing. The new lineup includes high-performance servers, high-speed array storage systems, and an expanded range of Hybrid AI services. These offerings are intended to help organizations leverage massive data volumes to accelerate scientific discovery, drive engineering advancements, and foster business innovation. The launch addresses the rising demand for infrastructure capable of handling continuous AI processing, autonomous agentic work, and increasingly complex scientific workloads.

The new portfolio integrates the latest Lenovo ThinkSystem servers powered by sixth-generation AMD EPYC 9006 SP7 processors with robust data storage and connectivity solutions. The lineup also features Lenovo Neptune liquid cooling technology. Lenovo is extending its AI Factory services to cover scales ranging from enterprise-level deployments to gigawatt-scale facilities. These capabilities provide the necessary computing power, data infrastructure, and expertise to help customers deploy, manage, and scale complex AI and HPC workloads. The infrastructure is designed to function effectively in on-premises data centers, hybrid environments, and colocation facilities.

Scott Petty, Vice President of the Infrastructure Solutions Group at Lenovo, stated that AI and high-performance computing are moving in the same direction to meet critical customer needs. The goal is to convert increasingly complex workloads into tangible results. This involves extracting maximum usable potential from every watt of energy and overcoming tightening power and cooling constraints. Lenovo is combining high-performance infrastructure, Neptune liquid cooling technology, and expanded services to help organizations achieve these goals at scale.

The new infrastructure aims to reduce bottlenecks and improve performance and value across the entire system. The Lenovo ThinkSystem SC755 server builds on more than a decade of liquid cooling expertise. It is a dual-socket server featuring Neptune technology that removes up to 100 percent of heat using water. This design delivers up to 40 percent higher sustained performance compared to air-cooled systems of the same class. The server reduces reliance on evaporative air conditioning and lowers water usage in large-scale high-performance environments, including HPC, agentic AI, and colocation settings. Lenovo Dynamic Partitioning technology allows customers to easily convert dual-socket systems into single-socket configurations and switch back, increasing resource allocation flexibility.

The Lenovo ThinkSystem SR755 server is developed for high-bandwidth data transfer tasks. It supports various industries, including large-scale virtualization, engineering, financial modeling, automated electronic circuit design, and agentic AI infrastructure. The server offers high computing density and memory bandwidth while using less energy and providing better value compared to dual-socket configurations. The Lenovo ThinkSystem DE6800 and DE6500 servers are designed for storage systems. Both models are built for scalability and performance, optimized for block workloads that require high data transfer rates and low latency. The DE6800 is suited for applications needing reliable storage, while the DE6500 offers high performance at a cost-effective price with simple installation and management.

Lenovo has also obtained certification for compatibility with the Cornelis CN5000 Fabric. This provides customers with an additional option for high-performance connectivity in demanding AI and HPC environments, helping to reduce bottlenecks in data-intensive workloads. Dan McNamara, Senior Vice President and General Manager of Compute and Enterprise AI at AMD, noted that organizations need infrastructure that delivers outstanding computing performance, energy efficiency, and scalability. The new Lenovo ThinkSystem platforms powered by sixth-generation AMD EPYC CPUs will help customers accelerate heavy workloads, from engineering and scientific research to AI inferencing and agentic AI, laying a more efficient and scalable foundation for continuous innovation.

Lenovo is expanding its service portfolio for both enterprise and gigawatt-scale Hybrid AI Factory deployments. New capabilities include GPU installation, configuration, and tuning to help customers build and scale Hybrid AI Factory environments for complex workloads. Support for multi-node operations on Red Hat AI Enterprise, NVIDIA AI Enterprise, and NVIDIA Run:ai provides greater flexibility in creating work environments tailored to specific needs. These services help customers move from infrastructure installation to practical AI capabilities more quickly.