GPU Gems: Programming Techniques, Tips and Tricks for Real-Time Graphics

Chapter 28. Graphics Pipeline Performance

Cem Cebenoyan NVIDIA

Overview

Over the past few years, the hardware-accelerated rendering pipeline has rapidly increased in complexity, bringing with it increasingly intricate and potentially confusing performance characteristics. Improving performance used to mean simply reducing the CPU cycles of the inner loops in your renderer; now it has become a cycle of determining bottlenecks and systematically attacking them. This loop of identification and optimization is fundamental to tuning a heterogeneous multiprocessor system; the driving idea is that a pipeline, by definition, is only as fast as its slowest stage. Thus, while premature and unfocused optimization in a single-processor system can lead to only ...