AI032
Programming Massively Parallel Processors: A Hands-on Approach
Modern GPU Features and Future Outlook
Learning Objectives
- Identify the specialized hardware units responsible for real-time ray tracing and AI acceleration.
- Evaluate the performance gains provided by Deep Learning Super Sampling (DLSS) and similar temporal upscaling techniques.
- Analyze the transition from monolithic GPU dies to Multi-Chip Module (MCM) architectures.
- Forecast the impact of Unified Memory and CXL on future heterogeneous compute environments.