Exploiting inter-sequence correlations for program behavior prediction

Prediction of program dynamic behaviors is fundamental to program optimizations, resource management, and architecture reconfigurations. Most existing predictors are based on locality of program behaviors, subject to some inherent limitations. In this paper, we revisit the design philosophy and systematically explore a second source of clues: statistical correlations between the behavior sequences of different program entities. Concentrated on loops, it examines the correlations' existence, strength, and values in enhancing the design of program behavior predictors. It creates the first taxonomy of program behavior sequence patterns. It develops a new form of predictors, named sequence predictors, to effectively translate the correlations into large-scope, proactive predictions of program behavior sequences. It demonstrates the usefulness of the prediction in dynamic version selection and loop importance estimation, showing 19% average speedup on a number of real-world utility applications. By taking scope and timing of behavior prediction as the first-order design objectives, the new approach overcomes limitations of existing program behavior predictors, opening up many new opportunities for runtime optimizations at various layers of computing.

[1]  Michael F. P. O'Boyle,et al.  Method-specific dynamic compilation using logistic regression , 2006, OOPSLA '06.

[2]  Anand Sivasubramaniam,et al.  Statistical profiling-based techniques for effective power provisioning in data centers , 2009, EuroSys '09.

[3]  Chen Ding,et al.  Program locality analysis using reuse distance , 2009, TOPL.

[4]  Alfred V. Aho,et al.  Compilers: Principles, Techniques, and Tools , 1986, Addison-Wesley series in computer science / World student series edition.

[5]  Eric R. Ziegel,et al.  The Elements of Statistical Learning , 2003, Technometrics.

[6]  F. Gabbay.  Speculative Execution based on Value Prediction Research Proposal towards the Degree of Doctor of Sciences , 1996 .

[7]  Chen Ding,et al.  Locality phase prediction , 2004, ASPLOS XI.

[8]  Feng Mao,et al.  Exploiting statistical correlations for proactive prediction of program behaviors , 2010, CGO '10.

[9]  Alexandra Fedorova,et al.  Addressing shared resource contention in multicore processors via scheduling , 2010, ASPLOS XV.

[10]  Yale N. Patt,et al.  A Comparison Of Dynamic Branch Predictors That Use Two Levels Of Branch History , 1993, Proceedings of the 20th Annual International Symposium on Computer Architecture.

[11]  Xipeng Shen,et al.  Does cache sharing on modern CMP matter to the performance of contemporary multithreaded programs? , 2010, PPoPP '10.

[12]  Lizy Kurian John,et al.  The Role of Return Value Prediction in Exploiting Speculative Method-Level Parallelism , 2003, J. Instr. Level Parallelism.

[13]  J. Larus.  Whole program paths , 1999, PLDI '99.

[14]  Martin Burtscher,et al.  Exploring last n value prediction , 1999, 1999 International Conference on Parallel Architectures and Compilation Techniques (Cat. No.PR00425).

[15]  Brad Calder,et al.  Phase tracking and prediction , 2003, ISCA '03.

[16]  James E. Smith,et al.  The predictability of data values , 1997, Proceedings of 30th Annual International Symposium on Microarchitecture.

[17]  Chris Lattner,et al.  LLVM: AN INFRASTRUCTURE FOR MULTI-STAGE OPTIMIZATION , 2000 .

[18]  Michael F. P. O'Boyle,et al.  Mapping parallelism to multi-cores: a machine learning based approach , 2009, PPoPP '09.

[19]  Wei-Chung Hsu,et al.  Dynamic Profile Driven Code Version Selection , 2007 .

[20]  Philip Jonathan,et al.  On the use of cross-validation to assess performance in multivariate prediction , 2000, Stat. Comput..

[21]  Yutao Zhong,et al.  Predicting whole-program locality through reuse distance analysis , 2003, PLDI.

[22]  Alfred V. Aho,et al.  Compilers: Principles, Techniques, and Tools (2nd Edition) , 2006 .

[23]  Xipeng Shen,et al.  An input-centric paradigm for program dynamic optimizations , 2010, OOPSLA.

[24]  Mikko H. Lipasti,et al.  Exceeding the dataflow limit via value prediction , 1996, Proceedings of the 29th Annual IEEE/ACM International Symposium on Microarchitecture. MICRO 29.

[25]  Martin Hirzel,et al.  Dynamic hot data stream prefetching for general-purpose programs , 2002, PLDI '02.

[26]  John L. Henning.  SPEC CPU2000: Measuring CPU Performance in the New Millennium , 2000, Computer.

[27]  Martin Burtscher,et al.  Prediction Outcome History-Based Confidence Estimation for Load Value Prediction , 1999, J. Instr. Level Parallelism.

[28]  Martin Burtscher,et al.  Hybrid Load-Value Predictors , 2002, IEEE Trans. Computers.

[29]  Koen De Bosschere,et al.  Differential FCM: increasing value prediction accuracy by improving table usage efficiency , 2001, Proceedings HPCA Seventh International Symposium on High-Performance Computer Architecture.