SPEC CPU2006 sensitivity to memory page sizes

Authors:
Wendy Korn;Moon S. Chang
Affiliations:
IBM;IBM
Venue:
ACM SIGARCH Computer Architecture News
Year:
2007

Citing 7
Cited 4

Characterizing the d-TLB behavior of SPEC CPU2000 benchmarks

SIGMETRICS '02 Proceedings of the 2002 ACM SIGMETRICS international conference on Measurement and modeling of computer systems
SPEC CPU2000: Measuring CPU Performance in the New Millennium

Computer
Practical, transparent operating system support for superpages

OSDI '02 Proceedings of the 5th symposium on Operating systems design and implementationCopyright restrictions prevent ACM from being able to make the PDFs for this conference available for downloading
POWER5 System microarchitecture

IBM Journal of Research and Development - POWER5 and packaging
Operating system exploitation of the POWER5 system

IBM Journal of Research and Development - POWER5 and packaging
SPEC CPU2006 benchmark descriptions

ACM SIGARCH Computer Architecture News
IBM Power5 Chip: A Dual-Core Multithreaded Processor

IEEE Micro

SPEC CPU2006 memory footprint

ACM SIGARCH Computer Architecture News
Generation, Validation and Analysis of SPEC CPU2006 Simulation Points Based on Branch, Memory and TLB Characteristics

Proceedings of the 2009 SPEC Benchmark Workshop on Computer Performance Evaluation and Benchmarking
Investigating the impact of code generation on performance characteristics of integer programs

Proceedings of the 2010 Workshop on Interaction between Compilers and Computer Architecture
TLB Improvements for Chip Multiprocessors: Inter-Core Cooperative Prefetchers and Shared Last-Level TLBs

ACM Transactions on Architecture and Code Optimization (TACO)

Quantified Score

Hi-index	0.00

Visualization

Abstract

SPEC CPU2006 is a compute-intensive industry standard benchmark suite published in August 2006. This paper characterizes the memory access behavior of SPEC CPU2006 running on IBM POWER5+ microprocessors. We measure the maximum and average memory usage of the benchmarks to validate SPEC's memory requirement criteria. This paper also analyzes how different page sizes affect the performance of the benchmarks. The experiment reveals that 64 KB and 16 MB pages improve the performance up to 46.9% and 50.9%, respectively.