The newly launched, most cutting-edge GPU server from Supermicro features eight NVIDIA H100 Tensor Core GPUs. The new high-end GPU server system will permit higher intake temperatures because of its innovative airflow design, which would lower a data center’s total Power Usage Effectiveness (PUE) while keeping the maximum performance profile.
Supermicro is adding this new Universal GPU server to its range of GPU servers, which would currently be the largest in the sector. The 4U, 5U, and new 8U 8GPU server are the three different Universal GPU server systems that Supermicro presently provides. Up to 400W, 350W, and higher Intel and AMD CPUs are supported by the Universal GPU platforms.
“Supermicro is leading the industry with an extremely flexible and high-performance GPU server, which features the powerful NVIDIA A100 and H100 GPU,” said Charles Liang, President and CEO at Supermicro. “This new server will support the next generation of CPUs and GPUs and is designed with maximum cooling capacity using the same chassis. We constantly look for innovative ways to deliver total IT Solutions to our growing customer base.”
Reduced Power Consumption and Noise
The 8U server is made for a variety of computationally demanding applications and is meant for data centers to operate at peak efficiency. High-performance NVIDIA H100 GPUs are included into the new 8U GPU system. Large data sets may be kept in memory with a maximum memory capacity of 8TB, speeding up the execution of HPC or AI training applications.
Additionally, the sophisticated architecture is built for GPU-to-GPU communication, which cuts down on the time needed for HPC simulations or AI training. Data may also be accessed directly by the GPUs using the NVIDIA GPUDirect Storage, greatly enhancing efficiency.
Innovative airflow design would lower fan speeds, reducing power consumption, total cost of ownership (TCO), and noise in the data center. Additionally, Supermicro’s server system supports standard OCP DC rack topologies as well as AC and DC Power.
