Ημερομηνία : 21/01/2021
Ομιλητής: Δρ. Διονύσιος Διαμαντόπουλος, ΙΒΜ Research, Zurich Lab
Bit. This foundational element originated from the combination of mathematics and Claude Shannon’s Theory of Information. Coupled with the 50-year legacy of Moore’s Law, which is aging fairly gracefully, the bit has propelled the digitization of our world. However, the cloud data explosion, further accelerated by AI applications, depends upon a continued exponential performance growth in computing power being fueled by computing innovation. Advancing from Narrow AI to Broad AI will encompass the unification of learning and reasoning through neuro-symbolic systems, resulting in a form of AI which will perform multiple tasks, operate across multiple domains, and will learn continuously. This AI will signal the necessity of the transition of the computing foundational element from bits to neurons. Looking further, the union of physics and information leads to the emergence of Quantum Information Theory and the development of the quantum bit -the qubit- forming the basis of quantum computers. The future of computing will look fundamentally different than it has in the past. It will not be based on more and cheaper bits alone, but rather, it will be built upon bits + neurons + qubits. This future will enable the next generation of intelligent mission-critical systems and accelerate the rate of science-driven discovery. Starting from a philosophical standpoint about the motivation of researching “What’s Next” in computing, in this talk I would like to share our vision for “bits + neurons + qubits”. We will further discuss the role of a specific form of bits, i.e. the bits processed by reconfigurable logic blocks, which form the foundational element of the Field Programmable Gate Arrays (FPGAs) and we will review how such devices can contribute to shaping our vision.
Dr. Dionysios Diamantopoulos is a Research Scientist at the Cloud and AI Systems Research department of IBM Research Europe – Zurich Lab. His research interests include energy-efficient reconfigurable systems, transprecision computing, near-memory computing, high-level synthesis, and heterogenous-accelerated systems for hybrid cloud and AI. He holds a Ph.D. in computer science and engineering from ECE/Technical University of Athens (2015), and a Diploma of Engineering from CEID/University of Patras (2009). His research is disseminated in more than 35 publications of peer-reviewed international conferences and journals. He has filed 4 patents with USPTO, for which he was awarded from IBM’s Council for Innovation Leadership. He has served in the TPC of DATE2020 conference and ATCET workshop of ISC18. He recently co-organized the “CloudFPGA DevOps” workshop at the FPL2020 conference. He has served as a reviewer in numerous conferences and journals, including DAC, DATE, FPL, ACM TACO, IEEE Access, and MICRO. He is a member of HiPEAC, IEEE, IEEE CAS, OpenPOWER, and Technical Chamber of Greece