A Hybrid Framework for Fast and Accurate GPU Performance Estimation through Source-Level Analysis and Trace-Based Simulation
Authors:
- Xiebing WangXWXiebing Wang
- Kai HuangKHKai Huang
- Alois KnollAKAlois Knoll
- Xuehai QianXQXuehai Qian
- 1 February 2019
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 506-518
Abstract
This paper proposes a hybrid framework for fast and accurate performance estimation of OpenCL kernels running on GPUs. The kernel execution flow is statically analyzed and thereupon the execution trace is generated via a loop-based bidirectional branch search. Then the trace is dynamically simulated to perform a dummy execution of the kernel to obtain the estimated time. The framework does not rely on profiling or measurement results which are used in conventional performance estimation techniques. Moreover, the lightweight trace-based simulation consumes much less time than a fine-grained GPU simulator. Our framework can accurately grasp the variation trend of the execution time in the design space and robustly predict the performance of the kernels across two generations of recent Nvidia GPU architectures. Experiments on four Commercial Off-The-Shelf (COTS) GPUs show that our framework can predict the runtime performance with average Mean Absolute Percentage Error (MAPE) of 17.04% and time consumption of a few seconds. We also demonstrate the practicability of our framework with a real-world application.Keywords
This publication has 43 references indexed in Scilit:
- GPGPU-MiniBench: Accelerating GPGPU Micro-Architecture SimulationIEEE Transactions on Computers, 2015
- GpuTejas: A parallel simulator for GPU architectures2014
- A Performance Model for GPUs with CachesIEEE Transactions on Parallel and Distributed Systems, 2014
- TBPoint: Reducing Simulation Time for Large-Scale GPGPU Kernels2014
- Accelerating GPGPU architecture simulation2013
- Break down GPU execution time with an analytical method2012
- Automatically Tuning Sparse Matrix-Vector Multiplication for GPU Architectures2010
- Auto-tuning 3-D FFT library for CUDA GPUs2009
- A bridging model for parallel computationCommunications of the ACM, 1990
- Parallelism in random access machines1978