Optimization of depth and shadow pass rendering in tile based architectures

A method of graphics processing, comprising:

• configuring a portion of a graphics processing unit (GPU) to include at least one depth processing block, the at least one depth processing block being associated with at least one depth buffer, the portion of the GPU corresponding to at least, one of a binning pipe or a geometry pipe;
• identifying one or more depth passes of each of a plurality of graphics workloads, the plurality of graphics workloads being associated with a plurality of frames;
• processing each of the one or more depth passes in the portion of the GPU including the at least one depth processing block, each of the one or more depth passes being processed by the at least one depth processing block, the one or more depth passes being associated with the at least one depth buffer;
• transmitting, upon processing each of the one or more depth passes, each of the one or more depth passes to a render GPU pipe or a rendering pipe of the GPU;
• combining, upon transmitting each of the one or more depth passes, each of the one or more depth passes with at least one other GPU operation;
• and
• generating, upon combining each of the one or more depth passes with the at least one other GPU operation, a final frame or a final render target based on the one or more depth passes and the at least one other GPU operation,
• wherein each of the one or more depth passes is processed concurrently with pixel shading and a render pass to improve GPU performance or processing speed, and
• wherein at least one of the binning pipe or the geometry pipe is associated with sorting one or more primitives into one or more tiles or one or more bins.

An apparatus for graphics processing, comprising:

• means for configuring a portion of a graphics processing unit (GPU) to include at least one depth processing block, the at least one depth processing block being associated with at least one depth buffer, the portion of the GPU corresponding to at least one of a binning pipe or a geometry pipe;
• means for identifying one or more depth passes of each of a plurality of graphics workloads, the plurality of graphics workloads being associated with a plurality of frames;
• means for processing each of the one or more depth passes in the portion of the GPU including the at least one depth processing block, each of the one or more depth passes being processed by the at least one depth processing block, the one or more depth passes being associated with the at least one depth buffer;
• and
• means for transmitting, upon processing each of the one or more depth passes, each of the one or more depth passes to a render GPU pipe or a rendering pipe of the GPU;
• means for combining, upon transmitting each of the one or more depth passes, each of the one or more depth passes with at least one other GPU operation;
• and
• means for generating, upon combining each of the one or more depth passes with the at least one other GPU operation, a final frame or a final render target based on the one or more depth passes and the at least one other GPU operation,
• wherein each of the one or more depth passes is processed concurrently with pixel shading and a render pass to improve GPU performance or processing speed, and
• wherein at least one of the binning pipe or the geometry pipe is associated with sorting one or more primitives into one or more tiles or one or more bins.

A computer-readable medium storing computer executable code for graphics processing, the code when executed by a processor causes the processor to:

• configure a portion of a graphics processing unit (GPU) to include at least one depth processing block, the at least one depth processing block being associated with at least one depth buffer, the portion of the GPU corresponding to at least one of a binning pipe or a geometry pipe;
• identify one or more depth passes of each of a plurality of graphics workloads, the plurality of graphics workloads being associated with a plurality of frames;
• process each of the one or more depth passes in the portion of the GPU including the at least one depth processing block, each of the one or more depth passes being processed by the at least one depth processing block, the one or more depth passes being associated with the at least one depth buffer;
• and
• transmit, upon processing each of the one or more depth passes, each of the one or more depth passes to a render GPU pipe or a rendering pipe of the GPU;
• combine, upon transmitting each of the one or more depth passes, each of the one or more depth passes with at least one other GPU operation;
• and
• generate, upon combining each of the one or more depth passes with the at least one other GPU operation, a final frame or a final render target based on the one or more depth passes and the at least one other GPU operation,
• wherein each of the one or more depth passes is processed concurrently with pixel shading and a render pass to improve GPU performance or processing speed, and
• wherein at least one of the binning pipe or the geometry pipe is associated with sorting one or more primitives into one or more tiles or one or more bins.