GIGABYTE Delivering HPC Solutions Powered by NVIDIA Tesla V100 32GB & P4 GPUs
May 30, 2018

 

Taipei, Taiwan 30th May 2018GIGABYTE is proud to be an exhibitor at GTC Taiwan today and tomorrow, where we will be showcasing the G481-S80, our latest high-performance NVIDIA GPU-powered server for AI training and deep learning and which will include an optional liquid cooling kit for even better performance possibilities.

 

During the conference NVIDIA will feature the Tesla V100 Tensor Core 32GB data center GPU. The Tesla V100 is NVIDIA’s most powerful data center GPU, suitable for AI training & deep learning, HPC, database, and graph analytics workloads. The V100 will now be available in a 32GB high-bandwidth memory configuration, which doubles the memory capacity of the previous V100 16GB model. Providing 2X the memory capacity can improve deep learning training performance for next-generation AI models by up to 50%. Supporting larger data models improves developer productivity, leading to faster AI insights, as well as allowing HPC applications to run larger simulations more efficiently than ever before.

 

The Tesla V100 32GB GPU is well suited for use with GIGABYTE’s G481-S80 (supporting up to 8 x Tesla V100 SXM2 form factor GPUs) or with the G481-HA0 (supporting up to 10 x Tesla V100 PCIe card form factor GPUs). These systems can generate massive parallel computing power to help drive AI breakthroughs. The G481-S80 is a member of the HGX-T1 class and the G481-HA0 is a member of SCX-E3 class of NVIDIA GPU-Accelerated Server Platforms.

 

GIGABYTE features an ever-expanding number of server platforms qualified for NVIDIA GPU products, not only including the Tesla V100 32GB and 16GB for deep learning and AI training, but also the Tesla P4, the world’s fastest deep learning inference GPU for scale-out servers to enable smart, responsive AI-based applications. The P4 can slash inference latency by up to 10X in any hyper-scale infrastructure while providing an incredible 40X better energy efficiency compared to CPUs, to unlock a new wave of AI services previously impossible due to latency limitations.

 

For the list of GIGABYTE’s NVIDIA GPU qualified products, please visit: https://www.nvidia.com/en-us/data-center/tesla/tesla-qualified-servers-catalog/

 

For more information on GIGABYTE’s server products, please visit: http://b2b.gigabyte.com

 

For more information on GTC Taiwan, please visit: https://www.nvidia.com/zh-tw/gtc/

{{AsideSubscribeInfo.buttonText}}