Method for performing shader occupancy for small primitives

Abstract

A GPU includes shader cores and a shader warp packer unit. The shader warp packer unit may receive a first primitive associated with a first partially covered quad, and a second primitive associated with a second partially covered quad. The shader warp packer unit may determine that the first partially covered quad and the second partially covered quad have non-overlapping coverage. The shader warp packer unit may pack the first partially covered quad and the second partially covered quad into a packed quad. The shader warp packer unit may send the packed quad to the shader cores. The first partially covered quad and the second partially covered quad may be spatially disjoint from each other. The shader cores may receive and process the packed quad with no loss of information relative to the shader cores individually processing the first partially covered quad and the second partially covered quad.

Description

[0001] Relevant application data [0002] This application claims the benefit of U.S. Provisional Application Serial No. 63 / 060,653, filed August 3, 2020, which is hereby incorporated by reference. technical field [0003] The present disclosure relates to graphics processing units (GPUs), and more particularly, to methods for performing shader occupancy on small primitives. Background technique [0004] Modern GPUs include a collection of programmable, highly parallel computing engines and a collection of various fixed-function units. Fixed function units may include texture address generation and filtering units, primitive clipping units, culling units, viewport transformation units, binning units, rasterization setup and rasterization units, depth comparison units, blending units, and / or others unit. GPUs can be used for graphics-intensive operations and / or compute-intensive workloads. [0005] Graphics data can flow through the GPU in a pipelined fashion, performing s...

AI Technical Summary

Problems solved by technology

However, this quadrilateral merging method is still lossy

Images