Registration Link: https://forms.office.com/r/WtxBDD2QWu
Workshop Objective: The workshop is intended to introduce accelerated computing and AI-ML on the GPUs. GPU computing will also be introduced. The latest GPU and CPU architectures and the best methods to program them will be presented. AI-ML will be introduced and ease of doing AI-ML with the help of containers will also be discussed. On the second day of the workshop people will get a chance to work on the GPUs in hands-on sessions for both the performance coding and the AI-ML.
Day1:
Morning (2.5 to 3 Hrs) (Speaker : Prathu)
- Introduction to GPU Compute.
- N-ways to do accelerated Computing.
- Latest GPU architecture A100, H100 and Grace.
- Programming considerations and Cuda features.
Afternoon (2.5 to 3 Hrs) (Speaker : Amit)
- Introduction of AI/ML through GPUs and why GPUs are critical for AI/ML
- NVIDIA AI stacks’ introduction for both Deep Learning and classical Machine-learning:
- Data preprocessing.
- Model training.
- Model Inferencing.
- Model deployment and MLOPS cycle.
Day 2:
- Working with NGC containers.
- Hands-on session for Accelerated computing.
- Hands-on session for AI-ML
Last Date of Registration: 26th Oct 2022
Resources:
- 1_Introduction to GPU Computing.pdf
- 2_Nsight_System.pdf
- 3_standards.pdf
- 4_OpenACC.pdf
- 5_OpenMP.pdf
- 6_CUDA.pdf
- AI Computing Workshop.pdf
- NeMo Getting Started
- IISC1.ipynb
- IISC2.ipynb
- hpc.zip