The ThinkSystem SR680a V3 is an air cooled, two-socket system, featuring the world’s most powerful GPUs for supercharging AI and HPC workloads with time-to-market support for NVIDIA HGX AI supercomputing platform, including H100, H200 and the all-new B200 architectures.

The ThinkSystem SR685a V3 is a 2-socket, 8U system accelerated for maximum compute with NVIDIA NVLink at 900GB/sec and support for NVIDIA H100 or H200 GPUs.

The ThinkSystem SR780a V3 is a powerful 2-socket, 5U direct water-cooled system tailored to handle compute-intensive AI Workloads in the Data Center. The system delivers maximum acceleration using 8 NVIDIA GPUs fully connected with high-speed NVLink interconnects.

With up to eight NVIDIA L40S GPUs, the SR675 V3 helps implement the next generation AI, metaverse simulations, and cognitive decisions at scale.

The SR650 V3 increases AI modeling capabilities with up to 8 single-width or 3 double-width GPUs for increased processing speed and performance.

The SR665 V3 supports up to 3x of the latest NVIDIA GPUs and provides the increased analytics speed necessary for AI and machine learning.

The ThinkAgile HX AI-Ready solutions with NVIDIA GPUs, are validated to fast-track time to value, accelerate AI inferencing, and cater to diverse customer needs across all business verticals.

This sever enables improved efficiency, performance, security, and TCO for AI workloads such as modeling and inferencing.