https://hgpu.org/?p=13139
A Self-Optimizing Framework for Developing Metrology Software on Massive Parallel Processor Architectures