Supermicro Unveils AI Solutions with AMD Instinct MI350 GPUs

Supermicro has introduced new AI solutions powered by AMD Instinct MI350 series GPUs, announced in a press release. These solutions are designed to deliver optimized performance and efficiency for large AI training models and high-speed inference workloads.

The new Supermicro H14 GPU solutions utilize the latest AMD CDNA architecture and offer significant memory capacity, with up to 2,304 TB HBM3e per 8-GPU server. This setup is aimed at enhancing computation speed and scalability for AI, inference, and training tasks.

Supermicro's offerings include both liquid-cooled and air-cooled GPU solutions, optimized for unmatched performance, scalability, and efficiency. The systems are built on Supermicro's proven building block architecture, allowing businesses to efficiently upgrade and scale their workloads.

The AMD Instinct MI350 series, built on the 4th generation AMD CDNA architecture, provides exceptional performance and energy efficiency for a wide range of workloads, from AI training and inference to complex scientific simulations. These new GPU solutions are available for order, offering flexibility and efficiency for cloud service providers and enterprises aiming to power AI at scale.