Filters








7 Hits in 4.0 sec

AVERAGE-CASE ANALYSIS OF ISOSPEED SCALABILITY OF PARALLEL COMPUTATIONS ON MULTIPROCESSORS

KEQIN LI, XIAN-HE SUN
2000 International journal of high speed computing  
We investigate the average-case speed and scalability of parallel algorithms executing on multiprocessors. Our performance metrics are average-speed and isospeed scalability.  ...  By modeling parallel algorithms on multiprocessors using task precedence graphs, we are mainly interested in the effects of synchronization overhead and load imbalance on the performance of parallel computations  ...  The framework of our analysis of the average-case performance of parallel computations is as follows.  ... 
doi:10.1142/s0129053300000072 fatcat:6xvzsrsw7fdarktyykidsobhey

Average-case analysis of isospeed scalability of parallel computations on multiprocessors

Keqin Li, Xian-He Sun
Proceedings 13th International Parallel Processing Symposium and 10th Symposium on Parallel and Distributed Processing. IPPS/SPDP 1999  
We investigate the average-case speed and scalability of parallel algorithms executing on multiprocessors. Our performance metrics are average-speed and isospeed scalability.  ...  By modeling parallel algorithms on multiprocessors using task precedence graphs, we are mainly interested in the effects of synchronization overhead and load imbalance on the performance of parallel computations  ...  The framework of our analysis of the average-case performance of parallel computations is as follows.  ... 
doi:10.1109/ipps.1999.760444 dblp:conf/ipps/LiS99 fatcat:yrat6nn2ufaw3eek72lxfvxk44

Software support for multiprocessor latency measurement and evaluation

Yong Yan, Xiaodong Zhang, Qian Ma
1997 IEEE Transactions on Software Engineering  
Parallel computing scalability evaluates the extent to which parallel programs and architectures can effectively utilize increasing numbers of processors.  ...  In this paper, we compare a group of existing scalability metrics and evaluation models with an experimental metric which uses network latency to measure and evaluate the scalability of parallel programs  ...  Part of the experiments were conducted on the KSR-1 machines at Cornell University and at the University of Washington.  ... 
doi:10.1109/32.581326 fatcat:s7ix6hur45gutjhnz276oof3wm

Integrated range comparison for data-parallel compilation systems

Xian-He Sun, M. Pantano, T. Fahringer
1999 IEEE Transactions on Parallel and Distributed Systems  
problem sizes via scalability and performance crossing point analysis.  ...  AbstractÐA major difficulty in restructuring compilation, and in parallel programming in general, is how to compare parallel performance over a range of system and problem sizes.  ...  His research focuses on software tools for parallel programs and multiprocessor architectures, in particular, parallelizing compilers, symbolic program, and performance analysis.  ... 
doi:10.1109/71.770134 fatcat:bsn66o3jfbaavntbxcrpumsz3a

Application and accuracy of the parallel diagonal dominant algorithm

Xian-He Sun
1995 Parallel Computing  
The Parallel Diagonal Dominant (PDD) algorithm is a highly efficient, ideally scalable tridiagonal solver. In this paper, a detailed study of the PDD algorithm is given.  ...  Implementation results show that the analysis gives a good bound on the relative error, and the algorithm is a good candidate for the emerging massively parallel machines.  ...  The emerging parallel computers build on such microprocessors are noted for their scalable architecture and massively parallel processing.  ... 
doi:10.1016/0167-8191(95)00018-j fatcat:uetdelwk4vcsphibgwmupyfxly

Workload modeling for parallel processing systems

G. Haring, G. Kotsis
MASCOTS '95. Proceedings of the Third International Workshop on Modeling, Analysis, and Simulation of Computer and Telecommunication Systems  
The analysis itself can either focus on the evaluation of the whole SUT, or can be oriented towards the evaluation of a certain part or component of the system, which is then called component under study  ...  For example, in modeling the execution time of a parallel program to be executed on a parallel system, the SUT would be both, the workload and the processing system, with the workload (i.e. the single  ...  of input data (best case scenario) or trying to nd typical combinations (average case scenario).  ... 
doi:10.1109/mascot.1995.378717 dblp:conf/mascots/HaringK95 fatcat:sdv4ribsn5gu3iokz35ltyrrhy

Minimizing communication overhead for matrix inversion algorithms on hypercubes

Xiaodong Wang, V.P. Roychowdhury
Proceedings of 9th International Parallel Processing Symposium  
Parallel Gauss-Jordan matrix inversion algorithms on the hypercube multiprocessors have been extensively studied in the literature.  ...  Rigorous ana1yt:ical and numerical performance analysis of our parallel algorithms are presented as well.  ...  Theorem. 10 The Isospeed Metric of Scalability In [30] the scalabilty of a parallel algorithm-machine combination is defined based on the isospeed metric.  ... 
doi:10.1109/ipps.1995.395969 dblp:conf/ipps/WangR95 fatcat:v3de6vaiejefrmoqvhhypd6foq