by Sanford Russell

Shanghai Jiao Tong University (SJTU) has big plans for GPU Computing, including plans to build the fastest supercomputer of any school in China.

In recognition of that, NVIDIA today recognized the school as the latest CUDA Center Excellence, our highest honor for organizations carrying out groundbreaking work using NVIDIA GPUs and CUDA technology. It joins just 15 other major institutions worldwide – Barcelona Supercomputing Center Harvard, Tokyo Tech and Stanford, among them – that have received the distinction for their GPU computing efforts.

Through their work, these organizations are helping academics and scientists deliver world-changing research on everything from climate research and renewable energy, to genomics and bio-medicine.

One of the few top universities in China with expertise in science, engineering and medicine, SJTU plans to tackle a number of GPU-related projects. Among them:

  • Fastest university supercomputer – It plans to build next year the fastest supercomputer of any Chinese university, using CPUs and NVIDIA Tesla GPUs. Once deployed, it will support a wide range of university research, training and educational programs.
  • CUDA/GPU computing courses – It plans to offer a wide range of CUDA-related courses, including the annual “Agile Many-core Software Development” course, as well as monthly HPC seminars.
  • Partnerships and outreach – SJTU will work with Shanghai Supercomputer Center  to migrate existing large-scale MPI applications to CUDA across large numbers of GPUs. It will also work with CAPS enterprise to build a GPU computing community in APAC. SJTU will also host a CUDA online resource portal for students and researchers in South China and around the world.
  • Research – GPU-accelerated research at SJTU includes simulating and researching fusion energy, improving the performance of airborne optical devices and improving detection and understanding of the dependency relationship between genes.

To help SJTU meet its goals, NVIDIA will award grants and donate a range of GPU computing equipment and technical support.