TY - GEN
T1 - Exploring thread and memory placement on NUMA architectures
T2 - 13th International Conference on High Performance Computing, HiPC 2006
AU - Antony, Joseph
AU - Janes, Pete P.
AU - Rendell, Alistair P.
PY - 2006
Y1 - 2006
N2 - Modern shared memory multiprocessor systems commonly have non-uniform memory access (NUMA) with asymmetric memory bandwidth and latency characteristics. Operating systems now provide application programmer interfaces allowing the user to perform specific thread and memory placement. To date, however, there have been relatively few detailed assessments of the importance of memory/thread placement for complex applications. This paper outlines a framework for performing memory and thread placement experiments on Solaris and Linux. Thread binding and location specific memory allocation and its verification is discussed and contrasted. Using the framework, the performance characteristics of serial versions of lmbench, Stream and various BLAS libraries (ATLAS, GOTO, ACML on Opteron/Linux and Sunperf on Opteron, UltraSPARC/Solaris) are measured on two different hardware platforms (UltraSPARC/FirePlane and Opteron/HyperTransport). A simple model describing performance as a function of memory distribution is proposed and assessed for both the Opteron and UltraSPARC.
AB - Modern shared memory multiprocessor systems commonly have non-uniform memory access (NUMA) with asymmetric memory bandwidth and latency characteristics. Operating systems now provide application programmer interfaces allowing the user to perform specific thread and memory placement. To date, however, there have been relatively few detailed assessments of the importance of memory/thread placement for complex applications. This paper outlines a framework for performing memory and thread placement experiments on Solaris and Linux. Thread binding and location specific memory allocation and its verification is discussed and contrasted. Using the framework, the performance characteristics of serial versions of lmbench, Stream and various BLAS libraries (ATLAS, GOTO, ACML on Opteron/Linux and Sunperf on Opteron, UltraSPARC/Solaris) are measured on two different hardware platforms (UltraSPARC/FirePlane and Opteron/HyperTransport). A simple model describing performance as a function of memory distribution is proposed and assessed for both the Opteron and UltraSPARC.
UR - http://www.scopus.com/inward/record.url?scp=65549134101&partnerID=8YFLogxK
U2 - 10.1007/11945918_35
DO - 10.1007/11945918_35
M3 - Conference contribution
SN - 354068039X
SN - 9783540680390
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 338
EP - 352
BT - High Performance Computing - HiPC 2006 - 13th International Conference Proceedings
Y2 - 18 December 2006 through 21 December 2006
ER -