2024-05-03 EuroCC2 Multi-GPU Programming Bootcamp (hbcp2s24)

Registration

Please apply via the www.openhackathons.org registration page with your official email address to prove your affiliation.
The final participants will be selected and informed after the registration deadline has passed.

Contents

The Multi-GPU Programming Bootcamp will be hosted virtually for two days on May 6 - 7. Scaling applications to multiple GPUs across multiple nodes requires one to be adept at programming models and optimization techniques, and proficient at performing root-cause analysis using in-depth profiling to identify and minimize bottlenecks. The Multi-GPU Programming Bootcamp covers step-by-step ways to improve application performance using cues from profilers and provides an understanding of the underlying technologies and communication topology needed to leverage powerful NVIDIA® libraries to extract more performance from the system.

This bootcamp is co-organized by the High-Performance Computing Center Stuttgart (HLRS), Jülich Supercomputing Centre (JSC), Leibniz Supercomputing Centre (LRZ), Vienna Scientific Cluster (VSC), RISE Research Institutes Of Sweden (RISE), Linköping University (LiU), OpenACC organization and NVIDIA for EuroCC Austria, EuroCC@GCS and EuroCC Sweden, all National Competence Centres for High-Performance Computing.

Event Format

This two-day bootcamp will be hosted online in the Central European Summer Time (CEST) zone. All communication will be done through Zoom, Slack, and email.

Prerequisites

• Experience with C/C++, NVIDIA CUDA®, MPI.

Hands-On

Attendees will be given access to a GPU cluster for the duration of the bootcamp.

Daily Agenda (Subject to Change):

All times are in Central European Summer Time (CEST).

Cluster Dry Run: Friday May 3, 2024: 11:00 AM - 12:00 PM

Day 1 Monday, May 6, 2024: 09:00 AM - 3:00 PM

• 09:00 AM - 09:05 AM: Welcome
• 09:05 AM - 09:30 AM: Introduction to Nsight Systems and GPU Computing (Lecture and Read-only Lab)
• 09:30 AM - 10:30 AM: Overview of a Single GPU Code
• 10:30 AM - 12:30 PM: Parallelizing a Single GPU Code using CUDA memcpy, Streams, and an Introduction to Single Node Multi-GPU (P2P) (Lecture and Lab)
• 12:30 PM - 01:00 PM: Lunch
• 01:00 PM - 03:00 PM: Introduction to Multi-Node Multi-GPU (GPU Direct) programming (Lecture and Lab)

Day 2 Tuesday, May 7, 2024: 09:00 AM - 01:30 PM

• 09:00 AM - 11:00 AM: Introduction to NVIDIA Collectives Communications Library (NCCL) (Lecture and Lab)
• 11:00 AM - 01:00 PM: Introduction to NVSHMEM (Lecture and Lab)
• 01:00 PM - 01:30 PM: Demo of the Nsight Systems Profiler Report for a Multi-Node Application

Language

English

Lecturers / Training Assistants

Lecturer: Nvidia (tba)

Training Assistants: HLRS, JSC, LiU, LRZ, RISE, VSC

Prices and Eligibility

The course is open and free of charge for participants from academia and industry from the Member States (MS) of the European Union (EU) and Associated Countries to the Horizon 2020 programme. Please make sure to register with your official email address to prove your affiliation.

Legal Notices

This course is offered in cooperation by HLRS, JSC, LiU, LRZ, RISE, VSC, OpenACC.org and NVIDIA. Registration is done via www.openhackathons.org hosted by OpenACC-Standard.org. Your registration data will be transferred to these partners. For legal notes see Privacy Policy.