Optimal and Heuristic Code Generation for Explicitly Parallel Processors
نویسندگان
چکیده
We describe our ongoing research within the EPICOpt project aimed at optimal and near-optimal code generation for explicitly parallel processors. On the one hand, this includes the application of techniques from the operations research domain to decrease the solver time for integer linear programming formulations of code generation problems. On the other hand, we want to investigate when and why established heuristics fail, and to develop efficient approximation algorithms and near-optimal techniques that remain computationally feasible even for large problems. To aid our investigation of the performance of various approaches to code generation, we are also investigating novel ways of efficient processor simulation using just-in-time compilation techniques.
منابع مشابه
A Message-Passing Distributed Memory Parallel Algorithm for a Dual-Code Thin Layer, Parabolized Navier-Stokes Solver
In this study, the results of parallelization of a 3-D dual code (Thin Layer, Parabolized Navier-Stokes solver) for solving supersonic turbulent flow around body and wing-body combinations are presented. As a serial code, TLNS solver is very time consuming and takes a large part of memory due to the iterative and lengthy computations. Also for complicated geometries, an exceeding number of grid...
متن کاملFinal report for Ceniit project 01.06 (2001–2007) Optimization Algorithms for Integrated Code Generation
In this project we developed integrative methods for generating high-quality code for embedded and instruction-level parallel processor architectures, including regular architectures such as superscalars and VLIW, as well as irregular architectures, such as clustered VLIW and DSP processors. Even though today’s DSP processors are more compiler friendly than a decade ago or than today’s network ...
متن کاملSimple Code Generation for special UDLs
This paper focuses on transforming sequential perfectly nested loops into their equivalent parallel form. A special category of FOR nested loops is the uniform dependence loops (UDLs), which yield efficient parallelization techniques. An automatic code generation tool for shared and distributed memory machines, has been developed in order to automatically parallelize these perfectly nested loop...
متن کاملOptimal Integrated VLIW Code Generation with Integer Linear Programming
We give an Integer Linear Programming (ILP) solution that fully integrates all steps of code generation, i.e. instruction selection, register allocation and instruction scheduling, on the basic block level for VLIW processors. In earlier work, we contributed a dynamic programming (DP) based method for optimal integrated code generation, implemented in our retargetable code generator OPTIMIST. I...
متن کاملMixed Large-Eddy Simulation Model for Turbulent Flows across Tube Bundles Using Parallel Coupled Multiblock NS Solver
In this study, turbulent flow around a tube bundle in non-orthogonal grid is simulated using the Large Eddy Simulation (LES) technique and parallelization of fully coupled Navier – Stokes (NS) equations. To model the small eddies, the Smagorinsky and a mixed model was used. This model represents the effect of dissipation and the grid-scale and subgrid-scale interactions. The fully coupled NS eq...
متن کامل