Reducing misspeculation overhead for module-level speculative execution

Authors:
Fredrik Warg;Per Stenstrom
Affiliations:
Chalmers University of Technology, Göteborg, Sweden;Chalmers University of Technology, Göteborg, Sweden
Venue:
Proceedings of the 2nd conference on Computing frontiers
Year:
2005

Citing 20
Cited 1

Multiscalar processors

ISCA '95 Proceedings of the 22nd annual international symposium on Computer architecture
Dynamic speculation and synchronization of data dependences

Proceedings of the 24th annual international symposium on Computer architecture
Data speculation support for a chip multiprocessor

Proceedings of the eighth international conference on Architectural support for programming languages and operating systems
Clustered speculative multithreaded processors

ICS '99 Proceedings of the 13th international conference on Supercomputing
A Chip-Multiprocessor Architecture with Speculative Multithreading

IEEE Transactions on Computers
Piranha: a scalable architecture based on single-chip multiprocessing

Proceedings of the 27th annual international symposium on Computer architecture
Multiplex: unifying conventional and speculative thread-level parallelism on a chip multiprocessor

ICS '01 Proceedings of the 15th international conference on Supercomputing
Enhancing software reliability with speculative threads

Proceedings of the 10th international conference on Architectural support for programming languages and operating systems
Limits on Speculative Module-Level Parallelism in Imperative and Object-Oriented Programs on CMP Platforms

Proceedings of the 2001 International Conference on Parallel Architectures and Compilation Techniques
The Potential for Using Thread-Level Data Speculation to Facilitate Automatic Parallelization

HPCA '98 Proceedings of the 4th International Symposium on High-Performance Computer Architecture
Exploiting Method-Level Parallelism in Single-Threaded Java Programs

PACT '98 Proceedings of the 1998 International Conference on Parallel Architectures and Compilation Techniques
In Search of Speculative Thread-Level Parallelism

PACT '99 Proceedings of the 1999 International Conference on Parallel Architectures and Compilation Techniques
Improving Speculative Thread-Level Parallelism Through Module Run-Length Prediction

IPDPS '03 Proceedings of the 17th International Symposium on Parallel and Distributed Processing
Using Incorrect Speculation to Prefetch Data in a Concurrent Multithreaded Processor

IPDPS '03 Proceedings of the 17th International Symposium on Parallel and Distributed Processing
Architecture of the Atlas Chip-Multiprocessor: Dynamically Parallelizing Irregular Applications

ICCD '99 Proceedings of the 1999 IEEE International Conference on Computer Design
A Quantitative Assessment of Thread-Level Speculation Techniques

IPDPS '00 Proceedings of the 14th International Symposium on Parallel and Distributed Processing
Eliminating Squashes Through Learning Cross-Thread Violations in Speculative Parallelization for Multiprocessors

HPCA '02 Proceedings of the 8th International Symposium on High-Performance Computer Architecture
Improving Value Communication for Thread-Level Speculation

HPCA '02 Proceedings of the 8th International Symposium on High-Performance Computer Architecture
The Superthreaded Architecture: Thread Pipelining with Run-Time Data Dependence Checking and Control Speculation

PACT '96 Proceedings of the 1996 Conference on Parallel Architectures and Compilation Techniques
SimICS/sun4m: a virtual workstation

ATEC '98 Proceedings of the annual conference on USENIX Annual Technical Conference

Mixed speculative multithreaded execution models

ACM Transactions on Architecture and Code Optimization (TACO)

Quantified Score

Hi-index	0.01

Visualization

Abstract

Thread-level speculative execution is a technique that makes it possible for a wider range of single-threaded applications to make use of the processing resources in a chip multiprocessor.We consider module-level speculation, i.e., speculative threads executing the code after a module (i.e., a procedure, function, or method) call. Unfortunately, previous studies have shown that indiscriminate module-level speculation results in significant overheads, mainly due to frequent misspeculations. In addition to hurting performance, excessive overhead is harmful from a resource usage and energy efficiency standpoint. We show that the overhead when spawning speculative threads for all module continuations is on average three times as big as the time spent on useful execution on our baseline 8-way chip multiprocessorIn this paper, we present and make a detailed evaluation of a technique that aims at reducing the overhead associated with misspeculations. History-based prediction is used in an attempt to prevent speculative threads from being spawned when they are expected to cause misspeculations. We find that the overhead can be reduced with a factor of six on average compared to indiscriminate speculation. The impact on speedup is small for most applications, but in several cases speedup is slightly improved.