CUDA Programming - High-Performance Computing with GPUs
Offered By: freeCodeCamp
Course Description
Overview
Master CUDA programming and harness the power of GPUs for high-performance computing and deep learning in this comprehensive 11-hour 55-minute course. Begin with an introduction to the deep learning ecosystem before diving into CUDA setup and a C/C++ review. Explore GPU architecture and learn to write your first CUDA kernels. Delve into the CUDA API and optimize matrix multiplication techniques. Discover Triton, a language for writing fast GPU code, and create PyTorch extensions. Apply your skills by implementing an MNIST multi-layer perceptron. Access accompanying code on GitHub, connect with the instructor on various platforms, and gain practical experience to accelerate your high-performance computing projects.
Syllabus
⌨️ Intro
⌨️ Chapter 1 Deep Learning Ecosystem
⌨️ Chapter 2 CUDA Setup
⌨️ Chapter 3 C/C++ Review
⌨️ Chapter 4 Intro to GPUs
⌨️ Chapter 5 Writing your First Kernels
⌨️ Chapter 6 CUDA API
⌨️ Chapter 7 Faster Matrix Multiplication
⌨️ Chapter 8 Triton
⌨️ Chapter 9 PyTorch Extensions
⌨️ Chapter 10 MNIST Multi-layer Perceptron
⌨️ Chapter 11 Next steps?
⌨️ Outro
Taught by
freeCodeCamp.org
Related Courses
First Order Optical System DesignUniversity of Colorado Boulder via Coursera Arithmetic Circuit Complexity
Indian Institute of Technology Kanpur via Swayam Introduction to Quantum Computing for Everyone
The University of Chicago via edX Dynamic Programming, Greedy Algorithms
University of Colorado Boulder via Coursera Linear Algebra
YouTube