Automatic tuning of whole applications using direct search and a performance-based transformation system

Qasem, Apan; Kennedy, Ken; Mellor-Crummey, John

doi:10.1007/s11227-006-7957-2

Cited by 38 publications

(29 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We perform tiling for a different class of applications. Qaseem et al have used complex modeling and a direct search to find transformation parameters [28]. Ren et al presented an auto-tuning framework for software-managed memory hierarchies [29].…”

Section: Related Workmentioning

confidence: 99%

Practical Loop Transformations for Tensor Contraction Expressions on Multi-level Memory Hierarchies

Krishnamoorthy

Agrawal

2011

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. Modern architectures are characterized by deeper levels of memory hierarchy, often explicitly addressable. Optimizing applications for such architectures requires careful management of the data movement across all these levels. In this paper, we focus on the problem of mapping tensor contractions to memory hierarchies with more than two levels, specifically addressing placement of memory allocation and data movement statements, choice of loop fusions, and tile size selection. Existing algorithms to find an integrated solution to this problem even for two-level memory hierarchies have been shown to be expensive. We improve upon this work by focusing on the first-order cost components, simplifying the analysis required and reducing the number of candidates to be evaluated. We have evaluated our framework on a cluster of GPUs. Using five candidate tensor contraction expressions, we show that fusion at multiple levels improves performance, and our framework is effective in determining profitable transformations.

show abstract

Section: Related Workmentioning

confidence: 99%

Practical Loop Transformations for Tensor Contraction Expressions on Multi-level Memory Hierarchies

Krishnamoorthy

Agrawal

2011

Lecture Notes in Computer Science

View full text Add to dashboard Cite

show abstract

“…Different search algorithms are used to search the parameter space -Genetic algorithms, Simulated Annealing, Pyramid search, Window search and Random search. Qasem et al [74,72] use a modified version of pattern-based direct search algorithm to explore the same search space. In addition, they provide compiler-based tuning for whole applications.…”

Section: Auto-tuning Frameworkmentioning

confidence: 99%

Tuning parallel applications in parallel

2009

View full text Add to dashboard Cite

Auto-tuning has recently received significant attention from the High Performance Computing community. Most auto-tuning approaches are specialized to work either on specific domains such as dense linear algebra and stencil computations, or only at certain stages of program execution such as compile time and runtime.Real scientific applications, however, demand a cohesive environment that can efficiently provide auto-tuning solutions at all stages of application development and deployment. Towards that end, we describe a unified end-to-end approach to autotuning scientific applications. Our system, Active Harmony, takes a search-based collaborative approach to auto-tuning. Application programmers, library writers and compilers collaborate to describe and export a set of performance related tunable parameters to the Active Harmony system. These parameters define a tun- Active Harmony supports runtime adaptive code-generation and tuning for parameters that require new code (e.g. unroll factors). Effectively, we merge traditional feedback directed optimization and just-in-time compilation. This feature also enables application developers to write applications once and have the autotuner adjust the application behavior automatically when run on new systems. We evaluated our system on multiple large-scale parallel applications and showed that our system can improve the execution time by up to 46% compared to the original version of the program.Finally, we believe that the success of any auto-tuning research depends on how effectively application developers, domain-experts and auto-tuners communicate and work together. To that end, we have developed and released a simple and extensible language that standardizes the parameter space representation. Using this language, developers and researchers can collaborate to export tunable parameters to the tuning frameworks. Relationships (e.g. ordering, dependencies, constraints, ranking) between tunable parameters and search-hints can also be expressed.

show abstract

“…Line (15)(16)(17)(18) of Figure 4 defines the collection of integer and boolean operations supported by POET. These operations include integer arithmetics (+,-,*), integer comparisons (<,<=,>,>=,==,!=), and boolean arithmetics (!, && and ||).…”

Section: Fig 6 Poet Definition Of Loop Blocking and Permutationmentioning

confidence: 99%

“…POET supports existing iterative compilation frameworks [12], [13], [16], [18], [9]. In particular, POET's explicit parameterization is designed to clearly separate analysis and code generation phases from the search phase.…”

Section: Related Workmentioning

confidence: 99%

“…Recent research has demonstrated that empirical tuning of application performance can significantly improve the effectiveness of compiler optimizations [23], [16], [12], [13], [18], [9], [27]. In these approaches, sensitive restructuring transformations are parameterized and dynamically re- <code Sequence pars=(s1,s2) > @s1@ @s2@ </code> <code Loop pars=(i,start,stop,step) > @init=(if start=="" then "" else (i "=" start )); test=(if stop=="" then "" else (i "<=" stop )); incr=(if step=="" then "" else (i "+=" step )); @for (@init@; @test@; @incr@) </code> <xform Unroll pars=(input,uloop) tune=(ur=16)> if (!…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation