By default, Unity’s runtime shader loading behavior is:
The benefit of this behavior is that there is no upfront GPU memory usage or load time for shader variants. The downside is that there can be visible stalls when a variant is used for the first time, because the graphics driver must create the shader program on the GPU and perform any additional work.
To avoid visible stalls at performance-intensive times, Unity can ask the graphics driver to create GPU representations of shader variants before they are first needed. This is called prewarming.
Warning: Check the notes on graphics API support before choosing how to perform prewarming. On modern graphics APIs such as DX12, Vulkan, and Metal, only the experimental ShaderWarmup API is fully supported, because it lets you specify a vertex format. Using the other methods can result in wasted work and GPU memory, without fixing the stalls.
You can perform prewarming in the following ways:
The profiler marker for Unity creating a representation of shader variant data to send to the GPU is
Shader.Parse. The profiler marker for uploading the shader program to the GPU, and waiting for the GPU to perform any required work, is
For information on using the Unity Profiler, see Profile your application.