The Pawsey Supercomputing Centre

Content

Australia’s first GPU Hackathon!

The first Southern Hemisphere GPU Hackathon will take place in Fremantle, Western Australia, April 16-20, 2018

Registrations are now open and will close on February 16th. The registration form is available at the Oak Ridge National Laboratory website.

 

Hackathon goal

The goal of each hackathon is for current or prospective user groups of large hybrid CPU-GPU systems to send teams of at least 3 developers along with either:

  1. A (potentially) scalable application that could benefit from GPU accelerators, or
  2. An application running on accelerators that need optimization.

There will be intensive mentoring during this 5-day hands-on workshop, with the goal that the teams leave with applications running on GPUs, or at least with a clear roadmap of how to get there.

Mentors come from national laboratories, supercomputing centres, universities and vendors, and besides having extensive experience in programming GPUs, many of them develop the GPU-capable compilers and help define standards such as OpenACC and OpenMP.

Target audience and format

We are looking for teams of 3-5 developers with a scalable** application to port to or optimize on a GPU accelerator. Collectively the team should know the application intimately. If application is a suite of apps, no more than two per team is allowed and a minimum of 2 people per app must attend. Space will be limited to 8 teams.

** node-to-node communication implemented, but don’t be discouraged to apply if your application is less than scalable. We are also looking for breadth of application areas.

Ok, so how can I attend?

The entry period will stay open for only 4 weeks, per event! See events below for deadlines. Selected teams will be notified approximately a week after the deadline closes.

For the full schedule and registration details please visit:

https://www.olcf.ornl.gov/training-event/2018-gpu-hackathons/

 

 

 

What applications are you targeting?

No application domain specifically. We hope to have open-source community codes in need of porting individual modules. This is a great opportunity for grad students and post-docs.

What will we use to program the GPUs?

For beginners, we recommend starting with OpenACC, but we are open to other GPU programming paradigms for those that have apps that already have some partial GPU port.

Participation Costs

Participation in the training event is free of charge. The meeting room and lunches, as well as access to the supercomputers throughout the event are offered by participating sites.

Mentors and learning materials introduced by the instructors are sponsored by participating sites the following partner organizations:

Participants are responsible for their travels to the meeting venue, their accommodation, 4 dinners during the week, and personal expenses.

Venue and bookings

We have chosen the Esplanade Hotel located in the port city of Fremantle, in Perth, Western Australia.

Participants are responsible for making their own travel arrangements to Fremantle.  You are also responsible for 4 dinners and your own personal expenses during the week.

You are encouraged to stay at the Esplanade Hotel as a discounted rate of AUD 190 (single occupancy) has been arranged, you could book through a Rydges pages made available especially for this event or contact to hotel directly. A full buffet breakfast is available from 20 AUD per person.  The rate is available from Saturday 14th April if required.

GPU Hackathon Student Scholarships will be available for the event.

Who can I contact for more information or questions?

Please contact Maciej Cytowski (maciej.cytowski@pawsey.org.au) for questions related to the GPU Hackathon content.

Background

General-purpose Graphics Processing Units (GPGPUs) potentially offer exceptionally high memory bandwidth and performance for a wide range of applications. The challenge in utilizing such accelerators has been the difficulty in programming them.  Today, these devices can be programmed with various available extensions to C++ and Fortran programming languages, like CUDA, OpenACC or OpenMP. There is also a wide range of numerical libraries which can make use of GPU acceleration.

Hackathons are five day intensive hands-on mentoring sessions. They are designed to help computational scientists port their applications to GPUs using libraries, OpenACC, CUDA and other tools.

They are currently lead by the Oak Ridge Leadership Computing Facility (OLCF) at the Oak Ridge National Laboratory (ORNL).