Designing the Next Generation of AI Systems Powered by NVIDIA: Accelerating AI Deployment Worldwide

Today, the Lenovo ThinkSystem SR675 V3 was featured on stage at SIGGRAPH as an NVIDIA OVX server that will soon include the newly announced NVIDIA L40S GPUs. These servers will help professionals worldwide advance AI and bring generative AI applications like intelligent chatbots, search, and summarization tools to users across industries.  

“Lenovo is the leading provider of omniverse and OVX in the cloud and on prem,” said Kirk Skaugen, President of Lenovo Infrastructure Solutions Group. “With the Lenovo ThinkSystem SR675 V3 powered by the new NVIDIA L40S GPU, Lenovo is further extending our omniverse leadership into generative AI, helping businesses implement next generation AI, immersive metaverse simulations, and cognitive decisions at scale.” 

The new NVIDIA L40S GPU is a powerful, universal data center GPU designed to deliver breakthrough multi-workload acceleration for AI inference and training, graphics, and video applications. The versatile capabilities of the L40S GPU make it the premier platform for accelerating AI workloads with  NVIDIA AI Enterprise, an enterprise-grade, end-to-end software platform for generative AI, and powering industrial digitalization with the NVIDIA Omniverse platform. 

“Combining the power of Lenovo and NVIDIA technologies will elevate generative AI and industrial digitalization to the next level,” said Bob Pette, vice president of professional visualization at NVIDIA. “Creative and professional teams can tap into the new L40S GPUs to accelerate their workloads and revolutionize how they interact with the world.” 

The L40S delivers breakthrough multi-workload acceleration for large language model (LLM) inference and retraining, graphics, and video applications. As the premier platform for multi-modal generative AI, the L40S GPU provides end-to-end acceleration to power the next generation of AI-enabled audio, speech, 2D, video, and 3D applications. The L40S will also be a key component for the next generation of NVIDIA Omniverse OVX 3.0 platforms delivering the highest fidelity and most accurate digital twins.   

Lenovo is well positioned for supporting the L40S with many servers in the line-up optimized for the new GPUs. As fully simulated digital replicas of entire buildings, factories, distribution centers, retail stores or cities, digital twins are combined with generative AI to improve business processes and design outcomes. Lenovo is extending its collaboration with NVIDIA on its NVIDIA OVX system from building and operating virtual worlds to generative AI, delivering powerful performance for AI workloads in the data center.  

The ThinkSystem SR675 V3, the leading platform supporting the OVX architecture, features up to eight?NVIDIA L40S GPUs?to maximize performance and deliver the most advanced ray tracing with?NVIDIA RTX-accelerated graphics and AI capabilities in a robust and scalable platform, allowing customers to explore and unlock the full potential of generative AI. The Lenovo ThinkSystem SR675 V3 is the most versatile, accelerated computing platform on the market with three server configurations in one, including support for NVIDIA HGX A100 4-GPU systems with NVIDIA NVLink technology and Lenovo Neptune hybrid liquid cooling, as well as 4- or 8-GPU configurations, featuring NVIDIA L40S, NVIDIA H100 PCIe 80GB or NVIDIA H100 NVL 94GB servers in a compact 3U footprint.   

Together, these advancements serve to simplify the rollout of AI, making it accessible to organizations of all sizes and enable transformative intelligence across all industries. This builds on Lenovo’s major investments in AI solutions and infrastructure and implementation of the technology across both offerings and operations.