[RCAC Workshop] NCCL and Distributed GPU Operations
📅 Date: December 1, 2025 ⏰ Time: 11:00 AM - 12:00 PM 💻 Location: Virtual 🏫 Instructor: Jacob Verburgt
Who Should Attend
Students and researchers with basic programming experience who are interested in learning how to effectively leverage multiple GPUs for scientific computing and machine learning. Familiarity with Python or C++ programming is recommended, but no prior GPU programming experience is required.
What You’ll Learn
NCCL and Distributed GPU Operations training will cover fundamentals of distributed training using Nvidia’s NCCL platform, as well as its application in AI. The foundations of NCCL, complete with examples in C++ will first be covered. This will be followed by distributed training demonstrations in PyTorch, specifically focusing on using NCCL in Distributed Data Parallel and Fully Sharded Data Parallel.
By the end of the session, you’ll... Understand distributed GPU operations and apply distributed GPU Operations into PyTorch-based machine learning workflows.
Level
Beginner to Intermediate
🔗 Register now: REGISTER HERE