On Wednesday and Thursday , August 14–15, 2019, NAS will host a two-day GPU Bootcamp featuring an introduction to GPU programming with OpenACC on the first day, followed by an introduction to CUDA, NVIDIA profiling tools, and machine learning and deep learning with python on the second day. There will be several opportunities for lab exercises including a half-day spent on a mini-application challenge. See the detailed agenda below.

The event will be held at the NAS facility in Mountain View, CA., and a Webex meeting will be available for remote attendees.

To register, please send your name and affiliation to: support@nas.nasa.gov.

NASA GPU Bootcamp Agenda

Note: All times are Pacific Time.

Wednesday, August 14 (Day 1)

Check-In (8:30 – 9:00 a.m.)

Part 1: Introduction to GPU Programming with OpenACC (9:00 – 11:00 a.m.)

      Introduction to GPU programming 

• What is a GPU and Why Should You Care?
• What is GPU Programming?
• Available Libraries, Programming Models, Platforms

      Introduction to OpenACC 

• What is OpenACC and Why Should You Care?
• Profile-driven Development
• First Steps with OpenACC
• Lab 1

      OpenACC Data Management 

• CPU and GPU Memories
• CUDA Unified (Managed) Memory
• OpenACC Data Management
• Lab 2

Break (11:00 – 11:30 a.m.)

Part 1, Continued (11:30 a.m. – 12:30 p.m.)

• Gangs, Workers, and Vectors Demystified 
• GPU Profiles
• Loop Optimizations
• Lab 3

Lunch (12:30 – 1:30 p.m.)

Part 2: Mini-Application Challenge (1:30 p.m. – 6:00 p.m.)

• Overview of the mini-application 
• Review steps to acceleration 
• Application challenge using AWS and Linux academy infrastructure 

Break (3:00 – 3:30 p.m.)

• Application challenge (continued)

Thursday, August 15 (Day 2)

Check-In (8:30 – 9:00 a.m.)

Part 1: Introduction to CUDA, NVIDIA profiling tools (9:00 a.m. – 12:30 p.m.)

• Introduction to CUDA   
• Multi-GPU/Multi-Node in OpenACC/CUDA 
• NVIDIA profiling tools 

Lunch (12:30 – 1:30 p.m.)

Part 2: Machine Learning and Deep Learning on GPUs  (1:30 p.m. – 5:00 p.m.)

• Machine Learning and Deep Learning on NVIDIA GPUs with Python 
• Deep Learning on Multiple GPUs 
• Open Q&A Session