Bilkent University
Department of Computer Engineering
S E M I N A R

 

High Performance Software Solutions for Exascale Systems

 

Asst. Prof. Dr. Didem Unat
Koç University

Parallel computer architecture has become increasingly complex both in terms of memory subsystems and computing units, resulting in great programming challenges to the application developers. While applications must be optimized to fully utilize a large and diverse pool of resources, they must be programmed under an elegant and simple code representation that allows for an easy code migration on the next generation supercomputers. Data locality abstractions and asynchronous execution are seen as the potential solutions towards alleviating the programming and scalability challenges. While programming abstractions provide performance portability and programmer’s productivity, asynchronous execution meets the high demand for performance and scalability of real-life applications. In this talk, I will present software solutions that I develop and co-develop to easy application development process in terms of (1) performance modeling and analysis, (2) programming model and programming abstractions, and (3) runtime system. This talk will also feature the ExaSAT performance modeling framework, TiDA a tiling library, and Perilla runtime system.

Bio: Didem Unat joined Koç University in September 2014 as a full time faculty. Previously she was at the Lawrence Berkeley National Laboratory and worked at the Exascale Combustion Co-design center. She is the recipient of the prestigious Luis Alvarez Fellowship in 2012 at the Berkeley Lab.

Her research interest lies primarily in the area of high performance computing, parallel programming models, compiler analysis, runtime systems, and performance modeling. She is currently working on designing and evaluating programming models for state-of-the-art computer architectures and leading the programming abstractions for data locality effort internationally through PADAL workshop series. She received her Ph.D under Prof. Scott B. Baden's research group at University of California-San Diego. In her thesis, she developed the Mint programming model and its source-to-source compiler to facilitate GPGPU programming. She holds a B.S in computer engineering from Boðaziçi University. On her spare time, she enjoys biking, hiking, and making mosaics.

 

DATE: 21 March 2017, Tuesday @ 13:40
PLACE: EA-409