Abstract: Approximate computing has evolved over the years as a fundamental method to improve performance and energy-efficiency in embedded and high-end systems, using algorithmic techniques that sacrifice numerical precision without necessarily compromising the quality of the result. In this talk we will explore approaches on how to use two forms of approximate computing, significance-driven execution and dynamic trans-precision arithmetic, to achieve both performance and better resilience in computing systems. We will also explore how approximate computing can be used to exploit emerging hardware features that trade hardware reliability with higher performance.

Short Bio: Dimistris Nikolopoulos is the Director of ECIT, the Global Research Institute of Electronics, Communication and Information Technology (ECIT) and a Professor in the School ofElectronics, Electrical Engineering and Computer Scienceat Queen's University Belfast. He is an expert in system software for scalable computing. He explores solutions to improve the performance, efficiency and reliability of servers, high-performance computing and distributed systems . He is the recipient of a Royal Society Wolfson Research Merit Award, the NSF CAREER Award, the DOE CAREER Award, the IBM Faculty Award, the SFI-DEL Investigator Award and Best Paper Awards from some of the premier IEEE and ACM conferences in my area, including SC, PPoPP, and IPDPS.