Understanding and Mitigating Multicore Performance Issues on theAMD Opteron Architecture

Levesque, John; Larkin, Jeff; Foster, Martyn; Glenski, Joe; Geissler, Garry; Whalen, S. P.; Waldecker, B.; Carter, Jonathan; Skinner, David; He, Helen; Wasserman, Harvey J.; Shalf, John; Shan, Hongzhang; Strohmaier, Erich

doi:10.2172/918496

Cited by 13 publications

(9 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Bilgir et al simulate Facebook workloads to look for energy and performance benefits in assigning the correct number of cores and mapping applications effectively across CMPs [6]. The works by Carter et al [9] and Levesque et al [30] evaluate whether increasing core counts on Cray machines will improve scientific applications' performance by estimating their memory bandwidth contention. Finally, Hood et al [19] and Jin et al [25] break down expected contention by class for different architectural platforms using microbenchmarks.…”

Section: Related Workmentioning

confidence: 99%

Measuring interference between live datacenter applications

Kambadur¹,

Moseley²,

Hank³

et al. 2012

2012 International Conference for High Performance Computing, Networking, Storage and Analysis

View full text Add to dashboard Cite

Abstract-Application interference is prevalent in datacenters due to contention over shared hardware resources. Unfortunately, understanding interference in live datacenters is more difficult than in controlled environments or on simpler architectures. Most approaches to mitigating interference rely on data that cannot be collected efficiently in a production environment. This work exposes eight specific complexities of live datacenters that constrain measurement of interference. It then introduces new, generic measurement techniques for analyzing interference in the face of these challenges and restrictions. We use the measurement techniques to conduct the first large-scale study of application interference in live production datacenter workloads. Data is measured across 1000 12-core Google servers observed to be running 1102 unique applications. Finally, our work identifies several opportunities to improve performance that use only the available data; these opportunities are applicable to any datacenter.

show abstract

Section: Related Workmentioning

confidence: 99%

Measuring interference between live datacenter applications

Kambadur¹,

Moseley²,

Hank³

et al. 2012

2012 International Conference for High Performance Computing, Networking, Storage and Analysis

View full text Add to dashboard Cite

show abstract

“…Other groups have looked at multicore systems [1], [2], [4]- [6], [11], [13], [17], [18], examining contention arising from a combination of shared resources in the memory hierarchy. Our approach is distinct in that it breaks contention down to its individual components-contention for shared cache, front-side bus, HyperTransport, QuickPath Interconnect, and contention on the memory bus.…”

Section: Introductionmentioning

confidence: 99%

Performance impact of resource contention in multicore systems

Hood

Jin²,

Mehrotra³

et al. 2010

2010 IEEE International Symposium on Parallel &Amp; Distributed Processing (IPDPS)

View full text Add to dashboard Cite

Resource sharing in commodity multicore processors can have a significant impact on the performance of production applications. In this paper we use a differential performance analysis methodology to quantify the costs of contention for resources in the memory hierarchy of several multicore processors used in high-end computers. In particular, by comparing runs that bind MPI processes to cores in different patterns, we can isolate the effects of resource sharing. We use this methodology to measure how such sharing affects the performance of four applications of interest to NASA-OVERFLOW, MITgcm, Cart3D, and NCC. We also use a subset of the HPCC benchmarks and hardware counter data to help interpret and validate our findings. We conduct our study on high-end computing platforms that use four different quadcore microprocessors-Intel Clovertown, Intel Harpertown, AMD Barcelona, and Intel Nehalem-EP. The results help further our understanding of the requirements these codes place on their production environments and also of each computer's ability to deliver performance.

show abstract

“…1 show that contention for memory plays a significant role on a program's execution time in multi-core machines. See also [12]. For the 12-core machine, when there is only one copy of the program in execution (i.e., concurrency level equal to one), the difference in execution time is 14.1 sec, which can be attributed to accessing 7 GB of main memory.…”

Section: Effects Of Memory Contention On Multi-core Computersmentioning

confidence: 99%

“…al. discuss the AMD chips and also discuss methods to collect performance data and use that data to analyze the performance impact of multi-core processors [12]. That work uses micro benchmarks and other benchmarks to explore compiler switches for software optimization.…”

Section: Related Workmentioning

confidence: 99%

Predicting the Effect of Memory Contention in Multi-Core Computers Using Analytic Performance Models

Bardhan

Menascé

2015

IEEE Trans. Comput.

View full text Add to dashboard Cite

Analyzing and predicting the performance of applications that run on multi-core computers is essential. This paper demonstrates experimentally that memory contention resulting from multiple cores accessing shared memory resources can become a significant component (i.e., over 50 percent) of an application's execution time. The paper develops single-and multi-class analytic performance models for predicting the effect of memory contention on a job's execution time. The models consider local and remote memory as in NUMA architectures. Model validation was done using a micro-benchmark and programs from HBench, UnixBench, and SPECCPU2006 running on machines with 4, 12, and 16 cores. The paper shows how to derive the model parameters and demonstrates that there is a significant difference in predicted values when memory contention is ignored. For example, a model that ignores memory contention predicts an average execution time about four times smaller than the experimental value for a concurrency level of 18 while the model with memory contention predicts a value that is 90 percent of the experimental value for the same concurrency level.

show abstract

Understanding and Mitigating Multicore Performance Issues on theAMD Opteron Architecture

Abstract: Over the past 15

Cited by 13 publications

References 3 publications

Measuring interference between live datacenter applications

Measuring interference between live datacenter applications

Performance impact of resource contention in multicore systems

Predicting the Effect of Memory Contention in Multi-Core Computers Using Analytic Performance Models

Contact Info

Product

Resources

About