Future geometry pipeline

Lately I’ve been interested in modernizing the geometry pipeline in the rendering engine, for example reducing the vertex shaders. It’s a well known fact that geometry shaders are not making efficient use of the GPU hardware, but similar issues could apply for vertex and tessellation shaders too, just because there is a need for pushing data through multiple pipeline stages. This can result in memory traffic and allocation bottlenecks between geometry and pixel stages.

Read More »

Bindless Descriptors

The Vulkan and DX12 graphics devices now support bindless descriptors in Wicked Engine. Earlier and in DX11, it was only possible to access textures, buffers (resource descriptors) and samplers in the shaders by binding them to specific slots. First, the binding model limitations will be described briefly, then the bindless model will be discussed.

Read More »

Tile-based optimization for post processing

One way to optimize heavy post processing shaders is to determine which parts of the screen could use a simpler version. The simplest form of this is use branching in the shader code to early exit or switch to a variant with reduced sample count or computations. This comes with a downside that even the parts where early exit occur, must allocate as many hardware resources (registers or groupshared memory) as the heaviest path. Also, branching in the shader code can be expensive when the branch condition is not uniform (for example: constant buffer values or SV_GroupID are uniform, but texture coordinates or SV_DispatchThreadID are not), because multiple threads can potentially branch differently and cause divergence in execution and additional instructions.

Read More »

Improved normal reconstruction from depth

In a 3D renderer, we might want to read the scene normal vectors at some point, for example post processes. We can write them out using MRT – multiple render target outputs from the object rendering shaders and write the surface normals to a texture. But that normal map texture usually contains normals that have normal map perturbation applied on them, which some algorithms don’t like. Algorithms that need to ray trace/ray march from the surface in multiple directions around the normal direction are not good fit with these normal mapped textures, because there is a disconnect between what is in the normal map and what is the real surface triangle normal.

Read More »

Thoughts on light culling: stream compaction vs flat bit arrays

I had my eyes set on the light culling using flat bit arrays technique for a long time now and finally decided to give it a try. Let me share my notes on why it seemed so interesting and why I replaced the stream compaction technique with this. I will describe both techniques and make comparisons between them. Wall of text incoming with occasional code pieces, in brain dump style.

Read More »

Simple job system using standard C++

After experimenting with the entity-component system this fall, I wanted to see how difficult it would be to put my other unused CPU cores to good use. I never really got into CPU multithreading seriously, so this is something new for me. The idea behind the entity-component system is both to make more efficient use of a single CPU by having similar data laid out linearly in memory (thus using cache prefetching when iterating), and also making it easier to write parallelized code (because data dependecies are more apparent). In this post I want to talk about the job system I came up with. It will not only make sense in the entity-component system, but generally it should perform well for any large data processing task. I wanted to remain in standard C++ (~11) realm, which I found that it is entirely possible. However, it can be extended with platform specific extensions if wanted. Let’s break this blog up into multiple parts:

Read More »