A programming interface for NUMA shared-memory clusters

被引：0
作者：
Dormanns, M [1 ]
Sprangers, W [1 ]
Ertl, H [1 ]
Bemmerl, T [1 ]
机构：
[1] RWTH Aachen, Lehrstuhl Betriebssyst, D-52056 Aachen, Germany
来源：
HIGH-PERFORMANCE COMPUTING AND NETWORKING | 1997年 / 1225卷
关键词：
parallel programming interface; shared memory parallelization; NUMA multiprocessor;
D O I：
暂无
中图分类号：
TP [自动化技术、计算机技术];
学科分类号：
0812 ;
摘要：
We describe a programming interface for parallel computing on NUMA (Non-Uniform Memory Access) shared memory machines. Although the interest in this architecture is rapidly growing and more and more hardware manufacturers offer products of this type, there is still a lack in parallelization support. We developed SMI, the Shared Memory Interface and implemented it as a library on an SCI-coupled cluster of workstations. It aims at providing sophisticated support to account for the NUMA performance characteristic and to allow a step-by-step parallelization. We show it's application to the parallelization of a sparse matrix computation.
引用
收藏
页码：698 / 707
页数：10
相关论文
共 24 条
[1]  
ABANDAH GA, 1996, CSETR27796 U MICH DE
[2]  
ADVE SC, 1995, SHARED MEMORY CONSIS
[3]  
[Anonymous], SCALABLE SHARED MEMO
[4]  
Bemmerl T., 1993, International Journal of High Speed Computing, V5, P595, DOI 10.1142/S0129053393000244
[5]  
CHANDRA R, 1994, P INT C SUP 94 JUL, P274
[6]  
CLARK R, 1996, SCI INTERCONNECT CHI
[7]  
*CONV COMP CORP, 1994, CONV EX ARCH
[8]  
*DOLP INT SOL, 1995, AS SPARC SBUS SCI CL
[9]  
DORMANNS M, 1997, P HIGH PERF COMP
[10]  
FALFASI B, 1994, P SUPERCOMPUTING
123