UPC Performance and Potential: A NPB Experimental Study

El‐Ghazawi, Tarek; Cantonnet, François

doi:10.1109/sc.2002.10034

Cited by 72 publications

(78 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In applications that have good data locality only a small proportion of shared accesses are remote. For instance, in the Sobel benchmark [23] only 1.6% of the shared accesses are remote when running with 2048 UPC threads.…”

Section: Combining Dynamic With Static Coalescingmentioning

confidence: 99%

Using shared-data localization to reduce the cost of inspector-execution in unified-parallel-C programs

et al. 2016

View full text Add to dashboard Cite

Programs written in the Unified Parallel C (UPC) language can access any location of the entire local and remote address space via read/write operations. However, UPC programs that contain fine-grained shared accesses can exhibit performance degradation. One solution is to use the inspector-executor technique to coalesce fine-grained shared accesses to larger remote access operations. A straightforward implementation of the inspectorexecutor transformation resultin excessive instrumentation that hinders performance.This paper addresses this issue and introduces various techniques that aim at reducing the generated instrumentation code: a shared-data localization transformation based on Constant-Stride Linear Memory Descriptors (CSLMADs) [32], the inlining of data locality checks and the usage of an index vector to aggregate the data. Finally, the paper introduces a lightweight * Corresponding author loop code motion transformation to privatize shared scalars that were propagated through the loop body. A performance evaluation, using up to 2048 cores of a POWER 775, explores the impact of each optimization and characterizes the overheads of UPC programs. It also shows that the presented optimizations increase performance of UPC programs up to 1.8× their UPC hand-optimized counterpart for applications with regular accesses and up to 6.3× for applications with irregular accesses.

show abstract

Section: Combining Dynamic With Static Coalescingmentioning

confidence: 99%

Using shared-data localization to reduce the cost of inspector-execution in unified-parallel-C programs

et al. 2016

View full text Add to dashboard Cite

show abstract

“…There is a number of known optimization techniques that improve the efficiency and performance of the UPC codes [4,7]. The following optimizations have been applied to the implementation of the UPCBLAS routines whenever possible:…”

Section: Upc Optimization Techniquesmentioning

confidence: 99%

“…In [4] El-Ghazawi and Cantonnet established, through an extensive evaluation of experimental results, that UPC can potentially perform at similar levels to those of MPI. Besides, the one-sided communications present in languages such as UPC were demonstrated to be able to obtain even better performance than the traditional two-sided communications [5].…”

Section: Introductionmentioning

confidence: 99%

UPCBLAS: a library for parallel matrix computations in Unified Parallel C

González-Domínguez

Martín

Taboada

et al. 2012

Concurrency and Computation

View full text Add to dashboard Cite

The popularity of Partitioned Global Address Space (PGAS) languages has increased during the last years thanks to their high programmability and performance through an efficient exploitation of data locality, especially on hierarchical architectures such as multicore clusters. This paper describes UP-CBLAS, a parallel numerical library for dense matrix computations using the PGAS Unified Parallel C (UPC) language. The routines developed in UPCBLAS are built on top of sequential BLAS functions and exploit the particularities of the PGAS paradigm, taking into account data locality in order to achieve a good performance. Furthermore, the routines implement other optimization techniques, several of them by automatically taking into account the hardware characteristics of the underlying systems on which they are executed. The library has been experimentally evaluated on a multicore supercomputer and compared to a message-passing based parallel numerical library, demonstrating good scalability and efficiency.

show abstract

“…Initial results are very promising and show that UPC application performance can be competitive with MPI [20], and offers improved usability and programmer productivity [21].…”

Section: Related Workmentioning

confidence: 99%

Scaling communication-intensive applications on BlueGene/P using one-sided communication and overlap

Nishtala

Hargrove

Bonachea

et al. 2009

2009 IEEE International Symposium on Parallel &Amp; Distributed Processing

View full text Add to dashboard Cite

In earlier work, we showed that the one-sided communication model found in PGAS languages (such as UPC) offers significant advantages in communication efficiency by decoupling data transfer from processor synchronization.We explore the use of the PGAS model on IBM BlueGene/P, an architecture that combines low-power, quad-core processors with extreme scalability. We demonstrate that the PGAS model, using a new port of the Berkeley UPC compiler and GASNet one-sided communication layer, outperforms two-sided (MPI) communication in both microbenchmarks and a case study of the communication-limited benchmark, NAS FT. We scale the benchmark up to 16,384 cores of the BlueGene/P and demonstrate that UPC consistently outperforms MPI by as much as 66% for some processor configurations and an average of 32%. In addition, the results demonstrate the scalability of the PGAS model and the Berkeley implementation of UPC, the viability of using it on machines with multicore nodes, and the effectiveness of the BG/P communication layer for supporting one-sided communication and PGAS languages.

show abstract

UPC Performance and Potential: A NPB Experimental Study

Cited by 72 publications

References 4 publications

Using shared-data localization to reduce the cost of inspector-execution in unified-parallel-C programs

Using shared-data localization to reduce the cost of inspector-execution in unified-parallel-C programs

UPCBLAS: a library for parallel matrix computations in Unified Parallel C

Scaling communication-intensive applications on BlueGene/P using one-sided communication and overlap

Contact Info

Product

Resources

About