Accelerating Machine Learning with CUDA
Instructor: Asoc. prof. Dr. sc. ing. Arnis Lektauers
Prerequisites: Basic C/C++ and Python competency
Duration: 2h
Format: Training webinar
The training webinar covers the theoretical and practical principles of massively parallel GPU computing with CUDA technology in the context of machine learning. Among to the overview of the CUDA architecture and programming model, the seminar will discuss the advanced aspects of machine learning acceleration in GPU hardware perspective.
Content
- Overview of CUDA architecture and programming model:
- GPU evolution;
- CUDA GPU architecture;
- Brief revise of CUDA programming model;
- Overview of CUDA memory hierarchy;
- Using CUDA on HPC cluster.
- Hardware accelerated machine learning:
- Machine learning acceleration in GPU hardware perspective;
- CUDA libraries for machine learning (cuBLAS, cuSOLVER, cuRAND, cuTENSOR, TensorRT);
- Multi-GPU and multi-node machine learning;
- Introduction to CUDA Deep Neural Network library (cuDNN);
- Overview of GPU-accelerated data analytics and machine learning suite RAPIDS.
Video recording from the webinar is available on CoE RAISE YouTube channel.
This webinar is created within CoE RAISE. More courses and training materials on artificial intelligence and HPC related applications can be found at the CoE RAISE training services platform.
