This year we are offering two special events the day before the workshop (March 5th). The events will also take place in the BioScience Research Collaborative building at Rice University (same venue as the main workshop).
If you plan to attend the tutorial, you will need to add this to your registration during registration.
OCCA: portability layer for many-core thread programming (2:00-3:30pm)
The OCCA API enables an experienced programmer who is comfortable with programming in OpenCL, CUDA, or OpenMP to write a single implementation of their compute kernels that can be treated at run time as any of these three languages. In this way the best performing threading model can be chosen at run time for almost all modern mainstream many-core processors.
Instructor: Tim Warburton
Hybrid Programming Challenges for CPU+GPU+Cluster Parallelism (4:00-5:30pm)
This mini-tutorial will introduce attendees to the fundamentals of programming for “hybrid” CPU+GPU+Cluster Parallelism. It will highlight the challenges of integrating constructs in the OpenMP, OpenCL, and MPI programming models, and illustrate new approaches that can address these challenges. Examples will be drawn from experiences gained in the Habanero Extreme Scale Software research project at Rice University.
Instructor: Zoran Budimlic
Networking Reception ($25.00):
Unlike previous years, this year the reception is open to all attendees (make sure you add networking reception when you register). Please join us for an informal meet and greet, which will include hors d’oeuvres and drinks.