Citations: Optimization within a Unified Transformation Framework

17 citations found. Retrieving documents...

W. Kelly, Optimization within a unified transformation framework, Tech. Rep. CSTR -3725, University of Maryland (1996).

Home/Search Document Details and Download Summary Related Articles Check

This paper is cited in the following contexts:

OCEANS - Optimising Compilers for Embedded Applications - Barreteau, Bodin.. (1998) (3 citations) (Correct)

....in [12] which uses GA based optimisation for an autoparallelising compiler. In the OCEANS GA, traditional restructuring transformation sequences such as tiling, loop permutation, loop distribution, loop fusion, loop skewing and statement reordering are represented as multi dimensional mappings [10]. GA optimisation initialises a population of mappings using a combination of randomised methods and conventional compiler techniques. Thus, a population represents a subset of the transformation space for a program. Mappings representing transformed programs having good performance (i.e. low ....

W. A. Kelly. Optimization within a Unified Transformation Framework. PhD thesis, Univ. of Maryland, 1996.

Constraint-Based Array Dependence Analysis - William Pugh, David Wonnacott (1998) (14 citations) (Correct)

....exists (see Pugh [1992] for details) We use the Omega test for memory based analysis because we need complete dependence relations, not just direction vectors, to perform our additional analysis. This additional information can also play an important role in program transformation [Pugh 1991; Kelly 1996]. 6.2.1 Blume s Range Test (for the Polaris compiler) The Range Test [Blume 1995] was developed to perform (memory based) data dependence analysis in the presence of nonlinear terms, as part of the Polaris compiler. The Range Test produces symbolic expressions giving the upper and lower bound on ....

....flows between loop iterations, and thus it can be used to determine the legality of array privatization. It does not produce a description of which iterations are involved in a value based dependence (which can be valuable for other optimizations, such as minimizing interprocessor communication [Kelly 1996]) and does not provide a general mechanism for deducing conditions that will disprove a dependence. 6.4.2 Barthou, Collard, and Feautrier s Fuzzy Array Data flow Analysis. Paul Feautrier s array data flow analysis [Feautrier 1991] and its extension to handle nonaffine terms [Barthou et al. ....

[Article contains additional citation context not shown here]

Kelly, W. A. 1996. Optimization within a unified transformation framework. Ph.D. thesis, Dept. of Computer Science, The University of Maryland.

Genetic Algorithm Optimised Compilation of Fortran for.. - Andrew Nisbet (Correct)

....strategy. Further evaluation is required to establish the full extent of applicability of the GAPS techniques. This will form the next phase of our continuing programme. 2 Background The GAPS approach (see Figure 1) has been implemented using the Unified Transformation Framework (UTF) [2] (developed as part of the Omega Project s [1] in the petit research prototype compiler. The UTF was chosen because it offers a more comprehensive set of transformations than the originally proposed software infrastructure, MARS [3] The UTF provides two simple mathematical abstractions, namely ....

....of loop and synchronisation costs. Here, no attempt was made to model cache miss overheads, and communication costs were not considered because partitioning decisions were controlled by petit . The model of loop overhead and the values assigned to parameters were based on the scheme presented in [2]. 2. measured execution time on two processors. Evaluation functions normally return a fitness value which GAs seek to maximise. In GAPS, we seek to minimise the predicted overheads or the measured execution time. Thus, highly fit individuals have low predicted overheads measured execution time. ....

W.A. Kelly, Optimization within a Unified Transformation Framework, Ph.D. Thesis, University of Maryland, USA, 1996.

GAPS: Iterative Feedback Directed Parallelisation Using Genetic .. - Andy Nisbet (1998) (2 citations) (Correct)

....times of code produced by GAPS, PFA, PETIT from the Omega Project are compared for an ADI 1024 Theta 1024 benchmark on an SGI Origin2000. On this benchmark, GAPS produces code having 21 Gamma 25 improvements in parallel execution time. 1 Introduction State of the art optimising compilers [1,2,4] utilise transformation sequences which attempt to minimise individual sources of overhead largely in isolation, since considering all overheads simultaneously is not practical. An infinite number of transformations can be applied, and determining the eventual total overhead for a particular ....

....and are therefore independent of target architecture and application programs. Hybrid techniques can guarantee that the best solution generated is no worse than the conventional techniques used to hybridise the GA. 3. 1 GAPS Infrastructure GAPS uses the Unified Transformation Framework (UTF) [4] from the Omega Project [3] The UTF provides two simple mathematical abstractions: time mappings and space mappings which respectively and concisely specify i) the restructuring transformations applied, and ii) the partition of computation for parallel execution. Single Program Multiple Data ....

W.A. Kelly (1996), Optimization within a Unified Transformation Framework, PhD. Thesis, University of Maryland, USA.

GAPS: Genetic Algorithm Optimised Parallelisation - Nisbet (1998) (2 citations) (Correct)

....benchmark on an SGI Origin 2000. On this benchmark GAPS produces code having 21 Gamma25 improvements in parallel execution time when measured execution time is used as the evaluation function. Keywords: Genetic Algorithm, Parallel, Compiler 1 Introduction State of the art optimising compilers [1,2,3] utilise transformation sequences which attempt to minimise individual sources of overhead largely in isolation, since considering all Supported by EPSRC project GR K82291, nisbeta cs.man.ac.uk. overheads simultaneously is not practical. An infinite number of transformations can be applied, ....

....2 describes related work. Section 3 discusses the basics concepts of GA optimisation [4] and the means by which GAs may be applied to the problem of compilation parallelisation. Section 4 provides an overview of the GAPS framework and discusses aspects of the Unified Transformation Framework [3] of the Omega Project [5] Section 5 describes the current GAPS configuration and presents results for this configuration in parallelising an ADI benchmark. Section 6 presents conclusions and future work. 2 Related Work [6] described a knowledge based parallelisation environment which attempted ....

W.A. Kelly (1996), Optimization within a Unified Transformation Framework, PhD. Thesis, University of Maryland, USA.

GAPS: A Compiler Framework for Genetic Algorithm (GA) Optimised .. - Andy Nisbet (Correct)

....to each statement and its associated iteration space. The hypothesis of GAPS is tested with an evaluation of the performance of SPMD code produced by GAPS on an SGI Origin 2000. Preliminary results indicate that this hypothesis is correct. Introduction State of the art optimising compilers [1,2,4] utilise transformation sequences which attempt to minimise individual sources of overhead largely in isolation, since considering all overheads simultaneously is not practical. An infinite number of transformations can be applied, and determining the eventual total overhead for a particular ....

....application program on a target architecture. Information logged during overhead prediction instrumented execution will be used by hybrid GAs to identify transformation (s) which can reduce program overheads. Software Infrastructure GAPS uses the Unified Transformation Framework (UTF) [4] from the Omega Project [3] The UTF provides simple mathematical abstractions : i) A timemapping T p : i 1 ; i m ] f 1 ; f n ]jC specifies the iteration reordering transformation applied to statement p, where the i 1 ; i m iteration variables represent ....

W.A. Kelly (1996), Optimization within a Unified Transformation Framework, PhD. Thesis, University of Maryland, USA.

GAPS: A Compiler Framework for Genetic Algorithm (GA) Optimised .. - Andy Nisbet (Correct)

....transformation applied to each statement and its associated iteration space. The hypothesis of GAPS is tested with an evaluation of the performance of SPMD code produced by GAPS for an ADI 1024x1024 benchmark on an SGI Origin 2000. 1 Introduction State of the art optimising compilers [1,2,4] utilise transformation sequences which attempt to minimise individual sources of overhead largely in isolation, since considering all overheads simultaneously is not practical. An infinite number of transformations can be applied, and determining the eventual total overhead for a particular ....

....reproduction uses one mutation and two crossover operators with equal and fixed selection probability. GAPS preserves elitism 5 by selecting individuals for deletion from the lower ranked half of the population. 3. 2 GAPS Infrastructure GAPS uses the Unified Transformation Framework (UTF) [4] from the Omega Project [3] The UTF provides two simple mathematical abstractions: time mappings and space mappings which respectively and concisely specify i) the restructuring transformations applied, and ii) the partition of computation for parallel execution. Single Program Multiple Data ....

W.A. Kelly (1996), Optimization within a Unified Transformation Framework, PhD. Thesis, University of Maryland, USA.

In Search of a Program Generator to Implement.. - Cohen, Donadio.. (2005) (Correct)

No context found.

W. Kelly, Optimization within a unified transformation framework, Tech. Rep. CSTR -3725, University of Maryland (1996).

Putting Polyhedral Loop Transformations to Work - Bastoul, Cohen, Girbal.. (2003) (2 citations) (Correct)