Lenovo has unveiled the powerful ThinkSystem SR685a V3 server, a cutting-edge solution crafted in collaboration with AMD to address the growing demands of AI workloads in various industries.
The AI supercomputer is a force to be reckoned with, sporting dual 4th Generation AMD EPYC processors with 1.5TB of HBM3 memory capacity and 1TB/s peak aggregate theoretical GPU I/O bandwidth performance.
Up to 16 NVMe SSDs
With support for up to 3TB of RAM and the flexibility to accommodate either AMD Instinct MI300X GPUs or Nvidia H100/H200/B100 GPUs, the ThinkSystem SR685a V3 is a versatile powerhouse for handling complex AI tasks.
This 8U2S rack server also boasts up to 10 PCIe Gen5 x16 FHHL adapters, hot-swappable NVMe SSDs, and redundancy features like multiple hot-swappable PSUs and efficient cooling mechanisms.
Designed for seamless operation, the server is managed by an XClarity Controller2 (XCC2) and is compatible with various operating systems, making it a comprehensive solution for AI applications across different sectors.
According to Forrest Norrod, from AMD, this collaboration with Lenovo signifies a significant step in accelerating AI transformation for enterprises, offering robust end-to-end solutions to meet the evolving computing needs.
With the ability to enhance AI workloads in sectors like financial services, the ThinkSystem SR685a V3 stands out for its capability to detect fraud, streamline KYC initiatives, and drive efficiency in various operational processes.
The Lenovo ThinkSystem SR685a V3 server is now available for order, with further pricing details obtainable upon request.
