Ahmet Akkaş; "Reliable Computing", April 4, 2007, 13:40, FENS G035
  • FENS
  • Ahmet Akkaş; "Reliable Computing", April 4, 2007, 13:40, FENS G035

You are here

Faculty of Engineering and Natural Sciences
FENS SEMINARS

Reliable Computing

Ahmet Akkaş

Computer speeds increase as a result of advances in VLSI technology and innovative computer design. Fast computers let programmers write numerically intensive programs, but computed results can be far from the true results due to the accumulation of errors in arithmetic operations. Therefore, accurate and reliable computations have become more important.

Interval arithmetic and extended precision (i.e., quadruple precision) arithmetic are two techniques for accurate and reliable computing. With interval arithmetic, each data value is represented by two floating-point numbers that correspond to the endpoints of an interval, such that the true result is guaranteed to lie on this interval. Since interval arithmetic represents ranges of numbers, it also provides the ability to solve problems that cannot be efficiently solved using floating-point arithmetic. Using quadruple precision arithmetic is another technique and it increases the accuracy and reliability of numerical computations by providing floating-point numbers that have more than twice the precision of double precision numbers.

In this talk, the performance of interval arithmetic is presented when interval arithmetic is supported by only software, by only the compiler, and by the compiler and hardware. Furthermore, it is shown that how quadruple precision arithmetic units are designed in dual-mode to provide hardware support for both quadruple precision and interval arithmetic operations. To evaluate the efficiency of different interval arithmetic implementations, an interval-enhanced compiler and superscalar processor simulator are used to obtain cycle accurate simulation results. Furthermore, hardware modifications to a superscalar architecture to support these enhancements and synthesis results for dual-mode quadruple precision units are presented

Bio: Ahmet Akkaş received the B.S. degree in Electrical Engineering from Gazi University, Turkey, in 1990. From 1990 to 1993, he worked as electronics engineer at the Third Air Force Reinforcement Maintenance Base in Turkey, where he also completed his military service. After that he obtained a fellowship for his graduate studies and went to USA in 1994. He received the M.S. and Ph.D. degrees in Computer Science from Lehigh University, USA, in 1996 and 2001, respectively. During his graduate studies, he worked as a summer intern in 1999 and 2000, at Agere Systems (formerly the Microelectronics Division of Lucent Technologies) in Allentown, Pennsylvania. After finishing his PhD, he joined Koç University in 2001. He is currently an assistant professor in the Computer Engineering Department at Koç University. His research interests include computer arithmetic, computer architecture, and reliable computing.

April 4, 2007, 13:40, FENS G035