Method for achieving large-scale high-performance Linpack testing benchmark for GPDSP
A high-performance, large-scale technology, used in the detection of faulty computer hardware, complex mathematical operations, etc., to solve problems such as the inability to obtain GPDSP performance test results
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment Construction
[0046] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.
[0047] The method for realizing the GPDSP-oriented large-scale high-performance Linpack test benchmark of the present invention is to solve dense linear equations in a block-by-block manner. In the solving process, small serial tasks and small matrix inversion and triangular equation solving tasks that are not suitable for DSP core vector processing array vectorization calculations, such tasks have relatively small calculations, and the calculation data storage mode is not suitable for efficient vectors Processing is handled by the CPU core. For regular calculation tasks (such as matrix and matrix multiplication), it is suitable for DSP core vector processing array vectorization calculation. The calculation amount of this type of task occupies the vast majority of Linpack calculation, and the DSP core is responsible for it. The calcu...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 