UCSD Runs New AI Supercomputer from @Xi 

@Xi Computers, manufacturer of high-performance AI high-performance computing servers and NVIDIA GPU Clusters, reports the operational deployment of a new Turnkey AI Supercomputer to the University of California San Diego (UCSD). 

This new UCSD Supercomputer consists of a mix of CPU and GPU nodes. There are 10 1U Xi WebRAIDer for a total of 400-Core Intel Xeon 4316 CPU cluster. One of the 10 systems sports an additional NVIDIA A10 GPU with 9216 CUDA Cores. Switches and network cables are also part of the contract to enable a full integration with the existing HPC infrastructure. @Xi Computers also provided the @XiCS, Supercomputer Cluster Management Solution software package, which includes a SLURM HPC Queueing System, Multiple Server Deployment Automation and Monitoring tools for CPUs and GPUs. The complete Supercomputer has been fully L11 and L12 tested and tagged prior delivery.

The Center of Excellence on Matter Under Extreme Conditions at UC San Diego under the direction of Professor Farhat Beg and research scientists Dr. Joohwan Kim, and Dr. Ozgur Culfa, will use this new Xi Supercomputer to increase the computational capabilities dedicated to advanced Plasma Physics and Thermo-Nuclear Fusion research. 

Sources: Press materials received from the company and additional information gleaned from the company’s website.

More XI Computer Coverage