Welcome to the MI300 Architecture training course


This course is delivered through a combination of presentations and reinforcement activities, like Multiple Choice Questions and hands-on exercises using Jupyter Notebooks. 

The estimated time to complete this course is 1 hour.


What you will learn

•    Understand the fundamentals of accelerator-based computing and heterogeneous CPU–GPU computing.
•    Identify the key architectural differences between Accelerated Processing Units (APUs) and discrete Graphics Processing Units (GPUs).
•    Compare CPU and GPU architectures, understand the AMD MI300 GPU design, and select the best architecture for your workload.
•    Evaluate the programming models used for a range of scientific computing and HPC workloads.


Prerequisites

•   Foundation in Bash
•   Foundation in Scalability
•   Foundation in Software Engineering
•   Foundation in Testing

Last modified: Friday, 17 July 2026, 3:54 PM