DMA transfer method for single instruction multithreading mode in gpdsp

Abstract

The invention discloses a single-instruction multi-thread mode oriented method for DMA (Director Memory Access) transmission in a GPDSP (General Purpose Digital Signal Processor). The method comprises: transferring data of SIMT (Single Instruction Multiple Thread) programs irregularly stored in an out-of-core storage space to a VM (Vector Memory) of a core by configuring a DMA transmission transaction once; and after the transfer, regularly storing the data in the VM, and enabling vector calculation parts to concurrently access to the data. The method is simple in principle, convenient to operate and configurable in data transmission amount, can efficiently supply data to the SIMT programs in a background operation manner, and not only better supports the execution of the SIMT programs but also greatly improves the operational performance of the GPDSP.

Description

technical field [0001] The present invention mainly relates to the field of direct memory access components (Director Memory Access, DMA) of a general purpose digital signal processor (General Purpose Digital Signal Processor, GPDSP), and particularly relates to a single instruction multiple thread (Single Instruction Multiple Threads, SIMT) mode. The DMA transfer method for data requirements is used to neatly move the data of the SIMT program stored irregularly in the out-of-core storage space to the in-kernel vector storage unit (Vector Memory, VM) through DMA transfer transactions, which ultimately facilitates the parallel execution of each vector unit. . Background technique [0002] Digital Signal Processor (DSP), as a typical embedded microprocessor, is widely used in embedded systems. It has powerful data processing capabilities, good programmability, flexible use and low power consumption. Its characteristics have brought great opportunities to the development of si...

AI Technical Summary

Problems solved by technology

[0004] General scientific computing requires high-performance DSP, but traditional DSP has the following disadvantages when used in scientific computing: 1) The bit width is small, which makes the calculation accuracy and addressing space insufficient

However, the current vector memory access operation only provides access to vector data with specific changing rules such as continuous address or equal address stride, which cannot meet the needs of multi-threaded parallel execution.

Images