Fundamentals of Accelerated Computing with OpenACC

8 hours of instruction

Find out how to write and configure code parallelization with OpenACC, optimize memory movements between the CPU and GPU accelerator, and apply the techniques to accelerate a CPU-only Laplace Heat Equation to achieve performance gains.

OBJECTIVES

1. Profile and optimize your CPU-only applications to identify hot spots for acceleration
2. Use OpenACC directives to GPU-accelerate your codebase
3. Optimize data movement between the CPU and GPU accelerator

PREREQUISITES

None

SYLLABUS & TOPICS COVERED

1. Introduction
   • Meet the instructor and create an account
2. Introduction To Parallel Programming
   • Introduction to parallelism
   • The goals of OpenACC
   • Basic parallelization of code using OpenACC
3. Profiling With Open ACC
   • Compiling sequential and OpenACC code
   • The importance of code profiling
   • Profiling sequential and OpenACC multicore code
   • Technical introduction to the code used in introductory modules
4. Introduction To Open ACC Directives
   • The Parallel directive and the Kernels directive
   • The Loop directive
5. GPU Programming With Open ACC
   • Definition of a GPU
   • Basic OpenACC data management
   • CUDA Unified Memory
   • Profiling GPU applications
6. Data Management With Open ACC
   • OpenACC data directive/clauses and structured data region
   • OpenACC unstructured data region and update directive
   • Data management with C/C++ Structs/Classes
7. Loop Optimizations With Open ACC
   • Seq/Auto clause, Independent clause, Reduction clause, Collapse clause, Tile clause
   • Gang, Worker, Vector

SOFTWARE REQUIREMENTS

Each participant will be provided with dedicated access to a fully configured, GPU-accelerated workstation in the cloud.