Nvidia has been taken up to supply the Tokyo Institute of Technology with GPUs that will eventually power a brand new AI supercomputer. The supercomputer will be fastest of its kind in all of Japan. This collaboration marks Nvidia’s shift from just a graphics processor maker to a wider role as a supplier for the rapidly growing AI market.
Nvidia will be supplying Tesla P100 GPUs to the Tokyo Insitute of Technology for the supercomputer. The GPUs use Pascal processing architecture and as such, are perfect for cluster creation. The cluster in this particular case, will be known as TSUBAME3.0, and which will replace TSUBAME2.5. The new cluster will bring along twice the performance capacity to the table. Interestingly, the system will deploy both, which will together generate a massive total of 64.3 petaflops worth of processing power.
The new computer will hopefully be in action by summer. Tokyo Institute of Technology is planning to deploy it for research and educational purposes. However, it could also allow private organizations to deploy the super smart AI powered supercomputer for their benefit — at a price of course. Once completed, the TSUBAME3.0 will be the fastest supercomputer of its kind in Japan.
Speaking on the topic, Satoshi Matsuoka, Professor, Ph.D., of the High Performance Computing Systems Group, GSIC, Tokyo Institute of Technology said:
As we run out the clock on Moore’s law, performance enhancements will increasingly be driven by improvements in data access times that come from faster storage media and networks, innovative data access approaches and the improvement of algorithms that interact with data subsystems.
Meanwhile, Data Direct Networks (DDN) has been selected as the strategic storage infrastructure provider for the new TSUBAME3.0 supercomputing system. This announcement was also made today, by the research team working on the supercomputer, along with DDN executives.
1 comment