Requirements for AI training and machine learning is growing exponentially. Supermicro provides the accelerated compute infrastructure for increasing the performance of AI training by 9 times over current products available.
Supermicro Infrastructure for Demanding AI/ML and HPC Environments
The Supermicro SYS-110D-16C-FRAN8TP is an Intel Xeon D system based on Ice Lake D. This is a 1U server designed with front I/O and plenty of networking.
Supermicro, a key player in the market for enterprise computing, storage and networking solutions, has unveiled its most advanced GPU server, incorporating eight NVIDIA H100 Tensor Core GPUs.
We're tearing down a $70,000, 3000W AMD AS-2114GT-DNR server by Supermicro that could be used in a supercomputer cluster. Wendell from Level1Techs brought one of these for us to look at in this tear-down. This server has two full systems with 6 total AMD Instinct MI210 GPUs and two EPYC CPUs.
We are looking at a server optimized for AI and machine learning. Supermicro has done a lot of work to cram as much as possible into 2114GT-DNR (2U2N) - a density optimized server. This is a really cool construction: there are two systems in this 2U chassis. The two redundant power supplies are 2,600W each and we'll see why we need so much power. It hosts six AMD MI210 Instinct GPUs and the dual Epyc processors. See the level of engineering Supermicro put into the design of this server.
