Chapter 9 GPU architectures and concepts

This is from Chapter 9 of Parallel and High Performance Computing, Robey and Zamora, Manning Publications, available at http://manning.com

The book may be obtained at http://www.manning.com/?a_aid=ParallelComputingRobey

Copyright 2019-2020 Robert Robey, Yuliana Zamora, and Manning Publications Emails: brobey@earthlink.net, yzamora215@gmail.com

See License.txt for licensing information.

Measuring the BabelStream GPU stream benchmark

Build with make cd BabelStream && make -f CUDA.make Run with ./cuda-stream OpenCL version cd BabelStream && make -f OpenCL.make Run with ./ocl-stream

Roofline performance model for GPUs

cd cs-roofline-toolkit/Empirical_Roofline_Tool-1.1.0 cp ../../config.V100_gpu Config Run with ./ert Config/config.V100_gpu View results with ps2pdf Results.config.Vega20_gpu/Run.001/roofline.ps xpdf roofline.pdf Radeon card cd cs-roofline-toolkit/Empirical_Roofline_Tool-1.1.0 cp ../../config.Vega20_gpu Config && ./ert Config/config.Vega20_gpu ps2pdf Results.config.V100_gpu/Run.001/roofline.ps xpdf roofline.ps

Mixbench performance tool

cd mixbench && make CUDA_INSTALL_PATH=nvcc/.. Run with ./mixbench-cuda-ro OpenCL version cd mixbench && make OPENCL_INSTALL_PATH=oclinfo/.. Run with ./mixbench-opencl-ro

PCI Bandwidth Benchmark (Book: listing 9.1 - 9.2)

Build with make cd PCI_Bandwidth_Benchmark && make Run with ./PCI_Bandwidth_Benchmark

Cloverleaf

./CloverLeaf_Ivybridge_batch.sh ./CloverLeaf_Skylake_batch.sh ./CloverLeaf_Power9_batch.sh

Power plot (Book: listing 9.3)

python power_plot.py