Computer Science

Colloquium Details

Towards a New Computer Architecture: Athena and the Concept of Cognitive Computing

Abstract

In this talk I will present a novel Multicore Architecture for Real-Time Hybrid Applications (MARTHA) with time-predictable execution, low computational latency, and high performance that meets the requirements for control, emulation and estimation of next-generation power electronics and smart grid systems. Our approach allows a large class of cyber-physical systems to be expressed as adaptive hybrid models executable on a single hardware platform. I will also introduce Athena, a general-purpose version of MARTHA that dynamically adapts and optimizes its execution behavior according to a set of high-level application goals (e.g., performance to power ratio, hard real-time constraints, reliability and security awareness). Athena represents a new, non-Von Neumann model of computation and provides a framework for future, cognitive, probabilistic, and adaptive computer system architectures. Using various machine learning and control theory techniques, it is able to: reason about the trade-off between the precision of results and the computational time; evaluate whether or not more time will yield better answers; and reason about “goodness” of the current solution or the computational progress in a given application execution stage. This new paradigm of probabilistic, approximate computing is particularly appealing to many important and resource intensive applications such as big data analytics, embedded systems, image and video processing, and real-time signal or data processing, where under some modalities, result accuracy can be traded-off for better performance, hard real-time support, and energy efficiency.

Biography

Michel Kinsy graduated from the Massachusetts Institute of Technology (MIT) with a Ph.D. in Electrical Engineering and Computer Science in June 2013. His doctoral work is one of the first to develop algorithms and hardware techniques to emulate and control large-scale power systems at the microsecond resolution. Part of this work is currently being commercialized. His research interests lie in the general area of computer architecture, with particular emphasis on: cognitive, high-performance many-core architectures, network-on-chip (NoC) routing, self-aware, adaptive multicore architectures, application of machine learning techniques to hardware execution, hard real-time large-scale distributed computer and cyber-physical systems. Michel is an MIT Presidential Fellow and holds an M.S. in Electrical Engineering and Computer Science from MIT, a B.S.E. in Computer Systems Engineering and a B.S. in Computer Science, both from Arizona State University.