Automatic vectorizing method for heterogeneous SIMD expansion components

Abstract

The invention relates to the field of high-performance computing automatic parallelization, in particular to an automatic vectorizing method for heterogeneous SIMD expansion components. The automatic vectorizing method is suitable for the heterogeneous SIMD expansion components with different vector quantity lengths and different vector quantity instruction sets, a set of virtual instruction sets are designed, and an input C and a Fortran program can be converted into an intermediate representation of virtual instructions under an automatic vectorizing unified framework. The virtual instruction sets are automatically converted into vectorizing codes for the heterogeneous SIMD expansion components through solving virtualization of the vector quantity lengths and solving virtualization of the instruction sets so that a programmer can be free from complex manual vectorizing coding work. The vectorizing method is combined with relative optimizing methods, vectorizing recognition is carried out from different granularities, mixing parallelism of a circulation level and a basic block level is explored to the greatest extent through conventional optimization and invocation point optimization, the redundancy optimization is carried out on generated codes through the analysis about striding data dependence of a basic block, and executing efficiency of a program is effectively improved.

Description

technical field [0001] The invention relates to the field of automatic parallelization of high-performance computing, in particular to an automatic vectorization method for heterogeneous SIMD expansion components. Background technique [0002] Human beings' endless demand for computing power makes parallel computing technology more and more people's attention. Generally speaking, the development of parallel hardware technology is much faster than the development of parallel software technology. SIMD expansion components provide hardware support for improving program performance. In order to give full play to the performance of SIMD expansion, it is necessary to fully explore the parallelism in the program and develop vectorized programs with good scalability. [0003] Nowadays, the scale of computing problems is huge and the amount of calculation is large, and it is difficult to manually write vectorized programs. Automatic vectorization technology analyzes the characteristi...

AI Technical Summary

Problems solved by technology

[0007] Among the above three typical vectorization methods, the traditional vectorization analyzes the loop layer by layer, and cannot be vectorized when the dependency is the inner loop carrying a dependency ring; the superword parallel vectorization has certain randomness in the packet generation, which may cause The final vectorization strategy is inconsistent with the ideal result; pattern matching vectorization is only a supplement to the above two methods

These methods can only generate vectorized code for a SIMD instruction set with a single vector length, and their portability and scalability are not strong, and have certain limitations.

Images