Analyzing the processor bottlenecks in SPEC CPU 2000

Joshua J. Yi; Ajay Joshi; Resit Sendag; Lieven Eeckhout; David J Lilja

Analyzing the processor bottlenecks in SPEC CPU 2000

Joshua J. Yi, Ajay Joshi, Resit Sendag, Lieven Eeckhout, David J Lilja

Electrical and Computer Engineering

Research output: Contribution to conference › Paper › peer-review

6 Scopus citations

Abstract

The performance of a processor is limited by the specific bottlenecks that a benchmark exposes while running on that processor. Since the quantification of these bottlenecks can be extremely time-consuming, our prior work proposed using the Plackett and Burman design as a statistically-rigorous, but time-efficient method of determining the processor's most significant performance bottlenecks. In this paper, we use the Plackett and Burman design to quantify the magnitude of the bottlenecks in the SPEC CPU 2000 benchmark suite from the viewpoints of both performance and energy consumption. We then use Principal Components Analysis, and hierarchical and K-means clustering algorithms to determine the similarity of the benchmarks based on their energy-delay production bottlenecks.

Original language	English (US)
State	Published - Dec 1 2006
Event	SPEC Benchmark Workshop 2006 - Austin, TX, United States Duration: Jan 23 2006 → Jan 23 2006

Other

Other	SPEC Benchmark Workshop 2006
Country/Territory	United States
City	Austin, TX
Period	1/23/06 → 1/23/06

Find It

Find It at U of M Libraries

Cite this

@conference{441b90c261b94624b6588ba55d5032d5,

title = "Analyzing the processor bottlenecks in SPEC CPU 2000",

abstract = "The performance of a processor is limited by the specific bottlenecks that a benchmark exposes while running on that processor. Since the quantification of these bottlenecks can be extremely time-consuming, our prior work proposed using the Plackett and Burman design as a statistically-rigorous, but time-efficient method of determining the processor's most significant performance bottlenecks. In this paper, we use the Plackett and Burman design to quantify the magnitude of the bottlenecks in the SPEC CPU 2000 benchmark suite from the viewpoints of both performance and energy consumption. We then use Principal Components Analysis, and hierarchical and K-means clustering algorithms to determine the similarity of the benchmarks based on their energy-delay production bottlenecks.",

author = "Yi, {Joshua J.} and Ajay Joshi and Resit Sendag and Lieven Eeckhout and Lilja, {David J}",

year = "2006",

month = dec,

day = "1",

language = "English (US)",

note = "SPEC Benchmark Workshop 2006 ; Conference date: 23-01-2006 Through 23-01-2006",

}

TY - CONF

T1 - Analyzing the processor bottlenecks in SPEC CPU 2000

AU - Yi, Joshua J.

AU - Joshi, Ajay

AU - Sendag, Resit

AU - Eeckhout, Lieven

AU - Lilja, David J

PY - 2006/12/1

Y1 - 2006/12/1

N2 - The performance of a processor is limited by the specific bottlenecks that a benchmark exposes while running on that processor. Since the quantification of these bottlenecks can be extremely time-consuming, our prior work proposed using the Plackett and Burman design as a statistically-rigorous, but time-efficient method of determining the processor's most significant performance bottlenecks. In this paper, we use the Plackett and Burman design to quantify the magnitude of the bottlenecks in the SPEC CPU 2000 benchmark suite from the viewpoints of both performance and energy consumption. We then use Principal Components Analysis, and hierarchical and K-means clustering algorithms to determine the similarity of the benchmarks based on their energy-delay production bottlenecks.

AB - The performance of a processor is limited by the specific bottlenecks that a benchmark exposes while running on that processor. Since the quantification of these bottlenecks can be extremely time-consuming, our prior work proposed using the Plackett and Burman design as a statistically-rigorous, but time-efficient method of determining the processor's most significant performance bottlenecks. In this paper, we use the Plackett and Burman design to quantify the magnitude of the bottlenecks in the SPEC CPU 2000 benchmark suite from the viewpoints of both performance and energy consumption. We then use Principal Components Analysis, and hierarchical and K-means clustering algorithms to determine the similarity of the benchmarks based on their energy-delay production bottlenecks.

UR - http://www.scopus.com/inward/record.url?scp=84883378949&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84883378949&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:84883378949

T2 - SPEC Benchmark Workshop 2006

Y2 - 23 January 2006 through 23 January 2006

ER -

Analyzing the processor bottlenecks in SPEC CPU 2000

Abstract

Other

Find It

Other files and links

Fingerprint

Cite this