Advancing the course to exascale computing, NVIDIA Corporation (NASDAQ:NVDA) reported that the “NVIDIA® Tesla®” AI supercomputing platform supports the best 13 measured systems on the latest Green500 list of the most energy-efficient HPC systems across the globe. All of these systems use NVIDIA Tesla ‘P100’ data center GPU accelerators.
The highlights
NVIDIA announced performance data showcasing that NVIDIA Tesla GPUs have enhanced performance for HPC applications by three times over the Kepler architecture launched two years ago. This considerably boosts performance past what would have been anticipated by Moore’s Law, even before it commenced slowing in recent years.
Moreover, NVIDIA reported that its Tesla-V100 GPU accelerators, which combine traditional HPC and AI applications on same platform, are estimated to offer the U.S. Department of Energy’s Summit supercomputer with remarkable floating point and AI performance when it is launched later this year.
Ian Buck, the GM of Accelerated Computing at NVIDIA, mentioned that researchers taking on the challenges are looking for a robust, unified computing architecture to benefit from HPC and the latest developments in AI. The company’s AI supercomputing platform offers one architecture for data and computational science, offering a combination of capabilities to boost the rate of innovation and unravel the unsolvable.
Professor Satoshi Matsuoka reported that with TSUBAME 3.0 supercomputer, their objective was to offer a single powerful platform for AI and HPC with finest energy efficiency as a leading Japanese national supercomputers. The most vital point is that they recorded this result with a best production machine of multi-petascale. In fact, NVIDIA Tesla ‘P100’ GPUs enabled them to excel on these objectives so they can offer this radical AI supercomputing platform to boost their education and scientific research of the country.
In the last trading session, the stock price of NVIDIA declined more than 1% to close the day at $144.56.
