GPU Work Graphs are an exciting new paradigm for graphics developers to explore. At its core, GPU Work Graphs enable a live shader kernel to dispatch new workloads on-demand, without needing to circle back around to the CPU first. At a high level, this could…
- … enable more direct methods of solving complex problems.
- … reduce memory constraints and improve cache utilization.
- … simplify inter-pass dependencies and barrier-induced complexity.
- … improve GPU thread saturation.
While this primer is in no way intended to describe all the things you can do with Work Graphs, it does enumerate many of the features and processes we have come to rely upon when testing and developing against Work Graphs within AMD. We’re eager to show you what we’ve been working on and look forward to seeing what the community does in this brave new frontier!