发明名称 |
Redistribution of generated geometric primitives |
摘要 |
One embodiment of the present invention sets forth a technique for redistributing geometric primitives generated by tessellation and geometry shaders for per-vertex by multiple graphics pipelines. Geometric primitives that are generated in a first processing stage are collected and redistributed more evenly and in smaller batches to the multiple graphics pipelines for vertex processing in a second processing stage. The smaller batches do not exceed the resource limits of a graphics pipeline and the per-vertex processing workloads of the graphics pipelines in the second stage are balanced. Therefore, the performance of the tessellation and geometry shaders is improved. |
申请公布号 |
US8917271(B2) |
申请公布日期 |
2014.12.23 |
申请号 |
US201012897578 |
申请日期 |
2010.10.04 |
申请人 |
NVIDIA Corporation |
发明人 |
Rhoades Johnny S.;Hakura Ziyad S.;Kilgariff Emmett M.;Kirkland Dale L.;Allison Cynthia Ann Edgeworth;Wurstner Karl M.;Abdalla Karim M. |
分类号 |
G06T15/50;G06T17/20;G06T15/00 |
主分类号 |
G06T15/50 |
代理机构 |
Patterson & Sheridan LLP. |
代理人 |
Patterson & Sheridan LLP. |
主权项 |
1. A method for distributing generated geometric primitives for processing by streaming multiprocessors, the method comprising:
executing a hull shader program by a first set of the streaming multiprocessors to process surface patches and generate tessellation data for each surface patch including tessellation level-of-detail values that define generated geometric primitives for each surface patch; generating tasks for processing the generated geometric primitives based on resource limitations of a streaming multiprocessor configured to execute one of the tasks and the level-of-detail values; and distributing the tasks to a second set of the streaming multiprocessors to process the tasks according to a domain shader program to produce processed vertices of the surface patches, wherein a first streaming multiprocessor included in the first set of the streaming multiprocessors is not also included in the second set of the streaming multiprocessors. |
地址 |
Santa Clara CA US |