Parallel data processing method and parallel processor for effectively eliminating data access delay

Abstract

The invention provides a parallel data processing method for effectively eliminating data access delay, and the method comprises the following steps: dividing data capable of being processed in parallel into a plurality of working groups, and setting parameters for each working group; starting multiple threads for processing, wherein each thread processes one work item; If data needing to be accessed to the external memory is encountered, storing the contexts, enabling the current thread to be in dormancy waiting, switching hardware to the next thread, and carrying out operation; after the data is returned, notifying the dormant waiting thread to enter switching waiting at any time, and once hardware resources are released, carrying out the recalculation immediately. The invention further provides a parallel processor. The invention has the beneficial effects that the task switching cost and the branch skipping cost can be well reduced, and the waste of computing resources and the consumption of energy are reduced.

Description

technical field [0001] The invention relates to a processor, in particular to a parallel data processing method and a parallel processor for effectively eliminating data access delay. Background technique [0002] At present, the task switching cost of the existing parallel data processing method is relatively high. For example, the Chinese patent No. 200910190339.1, a parallel processor and its thread processing method, discloses a MVP (Multi-thread Virtual Pipeline, multi-threaded virtual pipeline) processor, which can effectively deal with a large number of parallel tasks, and has a high cost performance. However, when there are relatively many data access instructions, and the computing instructions and the maximum number of threads cannot hide the delay, the parallel processor cannot adapt to the task characteristics well, resulting in high task switching costs, waste of computing resources and excessive consumption of energy. Contents of the invention [0003] In or...

AI Technical Summary

Problems solved by technology

However, when there are relatively many data access instructions, and the calculation instructions and the maximum number of threads cannot cover up the delay, the parallel processor cannot adapt to the task characteristics well, resulting in high task switching costs, waste of computing resources and excessive energy consumption

Images