Method for scheduling thread bundle warp, processor and computer storage medium
A technology of processors and thread bundles, applied in processor architecture/configuration, computing, program control design, etc., can solve problems such as waste of computing resources
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment Construction
[0023] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention.
[0024] As the basic scheduling unit for GPU to implement parallel execution, Warp width has a great impact on the performance and effect of parallel execution. Warp width usually indicates the number of threads contained in a warp. Currently, for GPUs, the warp width value is a fixed value. value, so a fixed-width warp cannot achieve the best performance results when facing the different needs of multiple applications. For example, when the warp width is 8, the best performance can be obtained in application scenario A; when the warp width is extended to 16, the best performance can be obtained in application scenario B. In the current GPU design, the warp width is fixed; if the warp width of a GPU is fixed to 8, the GPU can achieve the best performance in application scenario...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


