PERC 2 High-End Computer System Performance: Scalable Science and Engineering

Reed, Daniel

The University of North Carolina at Chapel Hill

关键词: 99 General And Miscellaneous//Mathematics, Computing, And Information Science;    Computers;    Parallel Processing;    Efficiency;    97 Mathematical Methods And Computing;   

DOI  :  10.2172/927308
RP-ID  :  DOE FC02-04ER25612
RP-ID  :  FC02-04ER25612
RP-ID  :  927308

美国|英语

来源: UNT Digital Library

【 摘 要 】

During two years of SciDAC PERC-2, our activities had centered largely on development of new performance analysis techniques to enable efficient use on systems containing thousands or tens of thousands of processors. In addition, we continued our application engagement efforts and utilized our tools to study the performance of various SciDAC applications on a variety of HPC platforms.