GPGPU Programming
Course Info
- Department: Mathematics and Computer Science, University of Calabria (Italy)
- Degree: Master Degree in Computer Science and Artificial Intelligence
- Curriculum: Data Science
- Lecturer: Donato D'Ambrosio
Course Material
Textbook
- David B. Kirk, Wen-mei W. Hwu, Programming Massively Parallel Processors: A Hands-on Approach (Third Edition). Morgan Kaufmann, 2017
Suggested readings
- CUDA Toolkit Documentation
- Grid-Stride Loops
- CUDA Unified Memory
- CUDA GPU Compute Capability
- Matching CUDA Architectures and Gencodes
- Query Device Properties and Handle Errors in CUDA
- CUDA Occupancy Calculator
- Unified L1/Texture Cache in Maxwell GPUs (see Section 1.4.2.1)
- Constant Memory (see Section 9.2.6)
- How to Implement Performance Metrics in CUDA C/C++
- Difference between dram_read_transactions and gld_transactions in CUDA profiler
Other resources
Slides
- Slides can be found in the Files section of the Lecture channel of the GPGPU Programming cours on Microsoft Teams
- You can join the GPGPU Programming team by using the following code: 3d98qzr