Upcoming Events

Location Application Deadline Event Date Team Proposals
Chinese Academy of Sciences, Huairou, Beijing, China  March 1, 2019 April 15-19, 2019 Closed
Universidad de los Andes, Colombia May 10, 2019 June 17-21, 2019 Apply Now
National Supercomputing Center in Shenzhen, China July 1, 2019 August 12-16, 2019 Apply Now
NCSA, Illinois, USA July 8, 2019 September 9-13, 2019 Apply Now
CDAC, Pune, India August 9, 2019 September 16-20, 2019  TBD

Why Participate

Participating in a hackathon provides a unique opportunity for teams to jumpstart acceleration or optimization of their code on GPUs with OpenACC or any other preferred tools. With the help of two experienced mentors, by the end of the event each team should have significantly accelerate their code or know which steps to take next to continue work.

Why Use GPUs

GPUs have thousands of cores and are proven to be powerful accelerators for parallel codes. Scientist and researchers see 2-10x performance increase while using GPUs for their applications. OpenACC is a directives-based programming model that is designed to help domain experts move their codes to GPUs faster. There is a variety of other tools available and we encourage you to explore the best one for your needs.

How to Prepare

  • Prior GPU knowledge is not required, but we encourage you to become familiar with OpenACC so that you can use your time at the hackathon more efficiently. A free OpenACC Online Course is available to everyone on openacc.org.

  • If your code is over 100,000 lines, please select and extract a stand-alone representative kernel from the original application to start working with at the event.

  • For more details on the structure of these events, please see the GPU Hackathon Attendee Guide, which is based on the experiences and suggestions of the organizers, mentors, and attendees over the years.

Costs and Prizes

Events are free for participants. Organizers will provide a room, lunches, mentors and access to compute resources. Hackathons are a collaboration and not a competition. The only prizes offered are new levels of performance for your code, time with experts, ability to run on supercomputers and unforgettable experience that might bring you closer to a new paper or talk.

Photography

Photographs of Participants may be taken during the OpenACC GPU Hackathons and be later published in various media by Organizer or Venue, without payment or other consideration to Participants, to share information pertaining to the OpenACC GPU Hackathons.

More GPU hackathons by Oak Ridge National Laboratory 

2019 Hackathon Call

What is your team name? Go ahead, be creative.
e.g. molecular dynamics, astrophysics, etc.
What does your application do? What is the size (LOC)? What language(s) and libraries are used in this app? Describe the current performance characteristics. Where does it run? (CPU, GPU) Describe what you envision in performance gains by porting to GPUs. Is it Open Source? Describe the size of your user community and the impact it might have. (Please answer all of the questions)
Is your model similar to cnn-resne50, lstm, BERT, random forest, etc? What optimizer and/or training method are you using? What systems have you worked before with your data and models? Please enter N/A if you don't do any machine learning or deep learning in your application.
Please add the size of a sample dataset that may be used for training purposes, and whether the data is labelled. Enter N/A if you don't use any machine learning or deep learning in your application.
Please specify the programming model or libraries you are planning to use for GPU acceleration. (e.g. CUDA, CUDA Fortran, OpenACC, OpenMP4.5+, cuBLAS, cuFFT, etc.)
Describe what sorts of algorithms dominate your application, especially the ones your team is targeting for acceleration.
Describe the current performance characteristics of your application. Where does it run (CPU, GPU)? How many nodes does it scale to?
What is the licensing of your application? *
Desktop, local clusters, HPC centers, etc.
Tell us about your team so we can better pair your team with experts. What is their familiarity with the application? What is their technical experience/background? Please add a statement on what you expect this hackathon will add to your teams experience.
We encourage teams to send as many team members as possible, including those early in their career and from underrepresented groups in HPC. A team between 3-6 members is ideal
List the following information about your team members, including you Order
Hackathom team members
With at least 2 people per app. If so, list the apps below along with the team members expected to work on them.

PHOTOGRAPHY

Photographs of Participants may be taken during the OpenACC GPU HACKATHONs and be later published in various media by Organizer or Venue, without payment or other consideration to Participants, to share information pertaining to the OpenACC GPU HACKATHONs.