There is a signup schedule for this event (UT EID required).

Type of T

alk:  UTCS Colloquium/Architecture

Speaker/Affiliation: 

Mike Kistler/IBM Austin Research Laboratory

Date/Time:  Monday

, April 27, 2009  3:30 p.m.

Location: ACES 2.302

Host:&n

bsp; Steve Keckler

Talk Title:  "Petascale Computing wit

h Accelerators"

Talk Abstract:
A trend is developing in hi

gh performance computing in which commodity processors are coupled to vario

us types of computational accelerators.  Such systems are commonly ca

lled hybrid systems. In this talk, I will describe our experience developi

ng an implementation of the Linpack benchmark for a petascale hybrid system

, the LANL Roadrunner cluster built by IBM for Los Alamos National Laborat

ory.  This system combines traditional x86-64 host processors with IB

M PowerXCell™ 8i accelerator processors.  The implementation of Linp

ack we developed was the first to achieve a performance result in excess of
1.0 PFLOPS, and made Roadrunner the #1 system on the Top500 list in June

2008. I will describe the design and implementation of hybrid Linpack, inc

luding the special optimizations we developed for this hybrid architecture.

  I will show actual results for single node and multi-node execution

s. From this work, we conclude that it is possible to achieve high perform

ance for certain applications on hybrid architectures when careful attentio

n is given to efficient use of memory bandwidth, scheduling of data moveme

nt between the host and accelerator memories, and proper distribution of w

ork between the host and accelerator processors.

Speaker Bio:
Mike Kistler is a Senior Technical Staff Member in the IBM Austin Resear

ch Laboratory. He received his BA in Math and Computer Science from Susqueh

anna University in 1982, an MS in Computer Science from Syracuse Universit

y in 1990, and MBA from Stern School of Business (NYU) in 1991. He joined

IBM in 1982 and has held technical and management positions in MVS, OS/2,
and Lotus Notes development. He joined the IBM Austin Research Laboratory

in May 2000 and is currently working on design and performance analysis for
IBM''s PowerPC and Cell/B.E. processors and systems. His research interest

s are parallel and cluster computing, fault tolerance, and full system si

mulation of high-performance computing systems.