Tom Gibbs is currently responsible for strategy and implementation of programs to enable and promote developers to take full advantage of NVIDIA technology. Tom brings over 25 years of experience in HPC, and has applications expertise in industries ranging from Aerospace to General Science, Healthcare, Life Sciences, Energy and Financial Services. Prior to NVIDIA Tom held senior management positions for early stage cloud startup companies in the healthcare market segment. He spent 15 years with the Intel Corporation, where he managed a global team responsible for leading innovation programs at CERN, NCSA, British Petroleum and Morgan Stanley as Director of Strategy and Architecture in the Solutions Group. During his time at Intel Tom was part of the HPC Business Unit responsible for ASCI RED and other large scale computing systems. Tom was a past Chairman of the Open Grid Forum and a member of the Center for Excellence in Supply Chain Management at MIT.
In the next 5 years three important factors are converging to increase in the pace and depth of scientific discovery. High Performance Computing (HPC) systems are evolving toward Exascale class, where throughput for is projected to increase by 50X for traditional simulation assisted science. In that time frame a new class of applications and workflows based on Deep Learning (DL) and Artificial Intelligence (AI) are emerging, where early examples highlight the potential to improve performance by 2 or more orders of magnitude with improved accuracy. Also, there are important new experiments and clinical systems that will increase the volume and resolution of data by 10X or more from sources that range from outer space, to subatomic particles, the human genome and cellular biology. We'll provide an historical overview of the opportunity and challenges we can expect to encounter to seize these new technologies, and then give multiple examples real world problems.