CUDA Courses and Certifications

Career Certificate

Introduction to CUDA Kernel Programming

Learn CUDA kernel programming from the ground up, including GPU memory, multidimensional kernels, shared-memory optimization, and image-processing pipelines.

• CodeSignal
• 9 hours
• Self-Paced
• Free Certificate
• 4 courses

CUDA Basics and 1D Operations

Learn CUDA fundamentals by allocating GPU memory, transferring data between CPU and GPU, and building reliable one-dimensional kernels with grid-stride loops.

• CodeSignal
• 3 hours
• Self-Paced
• Free Certificate

CUDA at Scale for the Enterprise

31 ratings at Coursera

Scale CUDA beyond a single GPU: coordinate multiple CPUs and GPUs, build asynchronous workflows with events and streams, and implement GPU sorting and image processing.

• Coursera
• 5 weeks, 6 hours a week
• Self-Paced
• Paid Course

CUDA Advanced Libraries

25 ratings at Coursera

Use CUDA Toolkit libraries — cuFFT, cuBLAS, cuSOLVER, Thrust, cuDNN and cuTensor — to accelerate Fourier transforms, linear system solving, sorting and neural networks on GPUs.

• Coursera
• 5 weeks, 5 hours a week
• Self-Paced
• Paid Course

Introduction to Parallel Programming with CUDA

Write CUDA programs for Nvidia GPUs: organize threads, blocks and grids, move data between host and global memory, and exploit shared, constant and register memory.

• Coursera
• 5 weeks, 4 hours a week
• Self-Paced
• Paid Course

Mastering Parallel programming with CUDA platform

1884 ratings at Udemy

Unofficial guide to parallel programming on GPU's with CUDA from basic concepts to advance algorithm implementations.

• Udemy
• 10 hours 48 minutes
• Self-Paced
• Paid Course

Packt

GPU Programming with C++ and CUDA

Master GPU programming in C++ and CUDA: write kernels, manage threads, blocks and streams, design parallel algorithms, profile with Nsight Compute, and expose GPU code to Python.

• Coursera
• 10 weeks, 1 hour a week
• Self-Paced
• Paid Course

CUDA Programming - High-Performance Computing with GPUs

3 ratings

Learn to write CUDA kernels and optimize GPU workloads, including matrix multiplication and integrations with Triton and PyTorch.

• freeCodeCamp
• 11 hours 55 minutes
• Self-Paced
• Free Video

Learn to use CUDA C/C++ tools and techniques to accelerate CPU-only applications to run on massively parallel GPUs.

• Independent
• 8 hours
• Self-Paced
• $90.00

CUTLASS: A CUDA C++ Template Library for Accelerating Deep Learning Computations

Explore CUTLASS, an open-source CUDA C++ template library for optimizing deep learning computations on NVIDIA GPUs. Learn to develop high-performance custom kernels using Tensor Core programming.

• YouTube
• 35 minutes
• Self-Paced
• Free Video

Nvidia

CUDA: New Features and Beyond

Discover the latest CUDA platform developments and future directions in this engineering-focused talk by a CUDA architect, exploring new features and the evolving GPU computing ecosystem.

• YouTube
• 44 minutes
• Self-Paced
• Free Video

Nvidia

1001 Ways to Write CUDA Kernels in Python

Explore best practices for writing CUDA kernels in Python, covering memory management, thread organization, and optimization techniques to harness GPU acceleration without leaving the Python ecosystem.

• YouTube
• 38 minutes
• Self-Paced
• Free Video

Nvidia

Unlocking GPU Performance with CUDA Tile

Discover how CUDA Tile revolutionizes GPU programming by abstracting tensor cores and simplifying kernel development for future-compatible, high-performance AI applications.

• YouTube
• 32 minutes
• Self-Paced
• Free Video

Nvidia

Accelerating Quantum Supercomputing with CUDA-Q

Unlock quantum supercomputing potential with CUDA-Q, an open-source platform for hybrid quantum-classical applications through expert panel discussions and practical guidance.

• YouTube
• 55 minutes
• Self-Paced
• Free Video

Nvidia

NVIDIA is a Software Company - CUDA - Episode 1

Dive into the world of CUDA with Stephen Jones, exploring parallelism, multicore compute history, and how 3D rendering concepts unlocked major machine learning algorithms.

• YouTube
• 1 hour 2 minutes
• Self-Paced
• Free Video