Benefits of speedup knowledge in memory-constrained multiprocessor scheduling

Parsons, Eric W.; Sevcik, Kenneth C.

doi:10.1016/s0166-5316(96)90030-9

Cited by 13 publications

(10 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several researchers [14]- [17] use application performance and concurrency characteristics such as speedup, execution time, synchronization information to make better scheduling decisions for parallel applications. Ferreira et al [18] showed how to quantify the application performance costs due to local OS interference on a range of real-world large-scale applications using over ten thousand nodes, and [19] identifies a major source of noise to be indirect overhead of periodic OS clock interrupts, that are used by all modern OS as a means of maintaining control.…”

Section: The Effect Of Os Level Factors and Scheduling Techniquesmentioning

confidence: 99%

Thread reinforcer: Dynamically determining number of threads via OS level monitoring

Pusukuri

Gupta

Bhuyan

2011

2011 IEEE International Symposium on Workload Characterization (IISWC)

View full text Add to dashboard Cite

Abstract-It is often assumed that to maximize the performance of a multithreaded application, the number of threads created should equal the number of cores. While this may be true for systems with four or eight cores, this is not true for systems with larger number of cores. Our experiments with PARSEC programs on a 24-core machine demonstrate this. Therefore, dynamically determining the appropriate number of threads for a multithreaded application is an important unsolved problem. In this paper we develop a simple technique for dynamically determining appropriate number of threads without recompiling the application or using complex compilation techniques or modifying Operating System policies. We first present a scalability study of eight programs from PARSEC conducted on a 24 core Dell PowerEdge R905 server running OpenSolaris.2009.06 for numbers of threads ranging from a few threads to 128 threads. Our study shows that not only does the maximum speedup achieved by these programs vary widely (from 3.6x to 21.9x), the number of threads that produce maximum speedups also vary widely (from 16 to 63 threads). By understanding the overall speedup behavior of these programs we identify the critical Operating System level factors that explain why the speedups vary with the number of threads. As an application of these observations, we develop a framework called "Thread Reinforcer" that dynamically monitors program's execution to search for the number of threads that are likely to yield best speedups. Thread Reinforcer identifies optimal or near optimal number of threads for most of the PARSEC programs studied and as well as for SPEC OMP and PBZIP2 programs.

show abstract

Section: The Effect Of Os Level Factors and Scheduling Techniquesmentioning

confidence: 99%

Thread reinforcer: Dynamically determining number of threads via OS level monitoring

Pusukuri

Gupta

Bhuyan

2011

2011 IEEE International Symposium on Workload Characterization (IISWC)

View full text Add to dashboard Cite

show abstract

“…Another solution to the memory problem guesses memory usage information based on information about the job provided by the user and information contained in the program executable [2]. Another uses speedup information known about jobs ahead of time to make scheduling decisions [4], [18]. Given this information, the scheduler can choose to give more processors to efficient jobs to increase utilization or to memory-intensive jobs to increase throughput.…”

Section: Previous Solutions To the Memory Problemmentioning

confidence: 99%

“…Performance will suffer from frequent page faults since hard disks are orders of magnitude slower than RAM. Research has shown that disk paging results in unsatisfactory performance on parallel platforms and should be avoided [2], [3], [4].…”

Section: Introductionmentioning

confidence: 99%

Effectively utilizing global cluster memory for large data-intensive parallel programs

Oleszkiewicz¹,

Xiao

2006

IEEE Trans. Parallel Distrib. Syst.

View full text Add to dashboard Cite

Abstract-Large scientific parallel applications demand large amounts of memory space. Current parallel computing platforms schedule jobs without fully knowing their memory requirements. This leads to uneven memory allocation in which some nodes are overloaded. This, in turn, leads to disk paging, which is extremely expensive in the context of scientific parallel computing. To solve this problem, we propose a new peer-to-peer solution called Parallel Network RAM. This approach avoids the use of disk, better utilizes available RAM resources, and will allow larger problems to be solved while reducing the computational, communication, and synchronization overhead typically involved in parallel applications. We proposed several different Parallel Network RAM designs and evaluated the performance of each under different conditions. We discovered that different designs are appropriate in different situations.

show abstract

“…The solution then uses this information in scheduling decisions [3]. If other information on jobs, such as speedup information, is known ahead of time, then another solution can use that information [15].…”

Section: Previous Solutions To the Memory Problemmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Parallel network RAM: effectively utilizing global cluster memory for large data-intensive parallel programs

Oleszkiewicz¹,

Xiao²

2004

International Conference on Parallel Processing, 2004. ICPP 2004.

View full text Add to dashboard Cite

show abstract

Benefits of speedup knowledge in memory-constrained multiprocessor scheduling

Cited by 13 publications

References 10 publications

Thread reinforcer: Dynamically determining number of threads via OS level monitoring

Thread reinforcer: Dynamically determining number of threads via OS level monitoring

Effectively utilizing global cluster memory for large data-intensive parallel programs

Parallel network RAM: effectively utilizing global cluster memory for large data-intensive parallel programs

Contact Info

Product

Resources

About