Supermicro Announces 8U ‘Universal GPU’ Server for NVIDIA H100’s 

Print Friendly, PDF & Email

Supermicro Universal GPU Servers

HPC-AI server maker Supermicro today announced what the company said is its most advanced GPU server incorporating eight NVIDIA H100 Tensor Core GPUs.

Supermicro now offers three Universal GPU servers: the 4U, 5U and the new 8U. The Universal GPU platforms also support Intel and AMD CPUs up to 400W, 350W and higher, according to the company.

“This new server will support the next generation of CPUs and GPUs and is designed with maximum cooling capacity using the same chassis,” said Charles Liang, president, and CEO of Supermicro. “We constantly look for innovative ways to deliver total IT Solutions to our growing customer base.”

With eight NVIDIA H100 GPUs, the system has a maximum memory capacity of 8TB, designed for faster execution of AI training or HPC applications. The architecture is designed for GPU-to-GPU communication, reducing the time for AI training or HPC simulations. In addition, NVIDIA GPUDirect Storage means “data can be directly accessed by the GPUs, further increasing efficiencies,” Supermicro said.

Supermicro said the system’s airflow design “will allow increased inlet temperatures, reducing a data center’s overall Power Usage Effectiveness (PUE) while maintaining the absolute highest performance profile.” The new airflow design “reduces fan speeds, resulting in less noise in the data center, lower power consumption, and a reduced total cost of ownership.”

The system also supports both AC and DC Power, including support for standard OCP DC rack configurations.

Supermicro said it supports open standards and adheres to the open power specification “assuring customers that Supermicro’s latest servers can be quickly delivered and installed at customer sites, resulting in faster time to productivity.”