HPC Achievement and Impact – 2018
TimeWednesday, June 27th5:30pm - 6:15pm
DescriptionOver the preceding year, the world of supercomputing has advanced at an unprecedented pace in terms of technologies, methodologies, and means. This presentation will review those changes, with regard to advances and challenges over the last 12 months – both from an individual perspective, strongly influenced by hardware, but also more broadly informed by application demands and system software developments. An emphasis on trends determining future directions and possibilities will be taken in the face of challenges imposed by the looming end of Moore’s Law, the race for exascale, the rapid growth of interest in machine learning and data analytics, and emerging paradigm shifts in computing methodologies. This year has seen coalescing schedules for achieving at least one typical metric of exascale across the beginning of the next decade employing a variety of architecture classes, even as the first systems to achieve 200 petaflops or more are being prepared before the end of this year. Machine learning is spawning multiple strategies for ever increasing productivity, enabling solutions to problems impossible only a few years ago. Quantum leaps in quantum computing have been pushed by both academia and industry, to the point that at least the glimmers of a shadow near the end of tunnel may be perceived. Other non-von Neumann architectures, such as neuromorphic computing, are being explored as well. There is much to discuss and many things to consider as the present status of the field of supercomputing presages its future impact.
This returning keynote is the 15th in its series and will be given before the closing presentation of ISC 2018.
Professor, Department of Intelligent Systems Engineering /Director, Center for Research in Extreme Scale Technologies