GPU Programming with OpenCL Training Course

Introduction

• What is OpenCL?
• Comparing OpenCL, CUDA, and SYCL
• Overview of OpenCL features and architecture
• Setting up the development environment

Getting Started

• Creating a new OpenCL project using Visual Studio Code
• Exploring project structure and file organisation
• Compiling and executing the program
• Displaying output using printf and fprintf

OpenCL API

• Understanding the role of the OpenCL API within host programs
• Querying device information and capabilities via the OpenCL API
• Creating contexts, command queues, buffers, kernels, and events using the API
• Enqueueing commands such as read, write, copy, map, unmap, execute, and wait
• Handling errors and exceptions through the OpenCL API

OpenCL C

• Understanding the role of OpenCL C in device-side programming
• Writing OpenCL C kernels to execute on the device and manipulate data
• Utilising OpenCL C data types, qualifiers, operators, and expressions
• Using OpenCL C built-in functions for math, geometry, and relations
• Leveraging OpenCL C extensions and libraries, including atomic, image, and cl_khr_fp16

OpenCL Memory Model

• Distinguishing between host and device memory models
• Using OpenCL memory spaces: global, local, constant, and private
• Managing OpenCL memory objects such as buffers, images, and pipes
• Applying OpenCL memory access modes like read-only, write-only, and read-write
• Utilising the OpenCL memory consistency model and synchronisation mechanisms

OpenCL Execution Model

• Distinguishing between host and device execution models
• Defining parallelism using OpenCL work-items, work-groups, and ND-ranges
• Using OpenCL work-item functions such as get_global_id, get_local_id, and get_group_id
• Employing OpenCL work-group functions like barrier, work_group_reduce, and work_group_scan
• Utilising OpenCL device functions such as get_num_groups, get_global_size, and get_local_size

Debugging

• Identifying common errors and bugs in OpenCL programs
• Using the Visual Studio Code debugger to inspect variables, set breakpoints, and analyse the call stack
• Debugging and analysing OpenCL programs on AMD devices using CodeXL
• Debugging and analysing OpenCL programs on Intel devices using Intel VTune
• Debugging and analysing OpenCL programs on NVIDIA devices using NVIDIA Nsight

Optimisation

• Understanding factors that influence OpenCL program performance
• Improving arithmetic throughput using OpenCL vector data types and vectorisation techniques
• Reducing control overhead and increasing locality through loop unrolling and loop tiling
• Optimising memory access and bandwidth using OpenCL local memory and its functions
• Measuring and improving execution time and resource utilisation through OpenCL profiling and tools