Nvidia Will Power Japan’s fastest AI Super Computer, Tsubame 3.0 Launching This Summer

Nvidia will partner with Tokyo Institute of technology for Japan’s fastest AI supercomputer known as Tsubame 3.0. Tsubame 3.0 is said to be much better than its predecessor Tsubame 2.5, Nvidia will provide the Pascal-based Tesla P1000 GPU technology, accelerating the performance in Tsubame 3.0.

Precisely, experts say that new fastest AI supercomputer: Tsubame 3.0 will be 3X efficient than its predecessor. This will be possible only with the latest GPU technology from Nvidia. Reportedly, new GPUs by Nvidia will deliver up to 12.2 petaflops of double precision performance.

Additionally, Nvidia says that a combination of Tsubame 3.0 and Tsubame 2.5 will pack a performance of 64.3 petaflops and this will make a combined system of Supercomputers to rank in top 10 around the globe.

Tokyo Tech’s Satoshi Matsuoka, a professor of computer science who is building the fastest AI super computer, further praised the partnership with Nvidia and said:

Nvidia’s broad AI ecosystem, including thousands of deep learning and inference applications, will enable Tokyo Tech to begin training Tsubame 3.0 immediately to help us more quickly solve some of the world’s once unsolvable problem

Tsubame 3.0

In parallel news, Nvidia’s top tier powerful yet reasonable GPU may be revealed soon at Nvidia’s GDC event. Nvidia’s GTX 1080 Ti is rumored to be in production and that it may arrive in late March.

The graphics card is expected to launch from March 20th to 23rd. The custom designed AIB versions of Nvidia’s GTX 1080Ti are expected to come out in the market sometime after the launch. Since the Geforce GTX 1080Ti will be slightly less in power than TITAN X, its price tag is expected to fall in the range of $750 to $900 at launch.

The graphics card is based on Nvidia’s GP102 Silicon. Nvidia will host its own show at GDC 2017 and we look forward to see and know more on GeForce GTX 1080Ti. To know more on this story, check out here.

Contributor at SegmentNext.