GPU Programming - Introduction to CUDA using C

Introduction to CUDA programming using C covers the fundamentals of parallel programming with NVIDIA’s CUDA platform, including concepts such as GPU architecture, memory management, kernel functions, and performance optimisation. The materials are designed for CUDA beginners and include step-by-step tutorials, practical examples, and exercises to help you get started with writing and running CUDA programs in C.

In this tutorial, we will learn how to write and compile basic CUDA programs, understand the execution model of CUDA-enabled GPUs, manage device and host memory, optimise code for better performance, and debug and profile CUDA applications.

1. Experience with C or C++ programming, especially heap memory.

2. Experience with using Gadi via command line interface.

• Write and compile basic CUDA programs.

• Understand the execution model of CUDA-enabled GPUs.

• Manage device and host memory.

• Optimize code for better performance.

The workshop content is subject to change without notice.

1. GPU Overview

2. GPU Execution Model

3. GPU Workflow

4. Asynchronous CUDA Calls

5. Shared Memory

6. CUDA Events

7. Unified Memory

The workshop content is subject to change without notice.