NaviSim: A Highly Accurate GPU Simulator for AMD RDNA GPUs

Abstract

As GPUs continue to grow in popularity for accelerating demanding applications, such as high-performance computing and machine learning, GPU architects need to deliver more powerful devices with updated instruction set architectures (ISAs) and new microarchitectural features. The introduction of the AMD RDNA architecture is one example where the GPU architecture was dramatically changed, modifying the underlying programming model, the core architecture, and the cache hierarchy. To date, no publicly-available simulator infrastructure can model the AMD RDNA GPU, preventing researchers from exploring new GPU designs based on the state-of-the-art RDNA architecture. In this paper, we present the NaviSim simulator, the frst cyclelevel GPU simulator framework that models AMD RDNA GPUs. NaviSim faithfully emulates the new RDNA ISA. We extensively tune and validate NaviSim using several microbenchmarks and 10 full workloads. Our evaluation shows that NaviSim can accurately model the GPU's kernel execution time, achieving similar performance to hardware execution within 9.92% (on average), as measured on an AMD RX 5500 XT GPU and an AMD Radeon Pro W6800 GPU. To demonstrate the full utility of the NaviSim simulator, we carry out a performance study of the impact of individual RDNA features, attempting to understand better the design decisions behind these features. We carry out a number of experiments to isolate each RDNA feature and evaluate its impact on overall performance, as well as demonstrate the usability and flexibility of NaviSim.