NVIDIA Workshop GPU Hackathon 2017
06 February, 2017
The 4th Annual Workshop on Accelerating Scientific Applications Using GPUs
was hosted by the KAUST Supercomputing Laboratory
. During the following two-day GPU hackathon event few selected teams of developers had the chance to port and accelerate their domain science application to GPU accelerators guided by OpenACC and CUDA mentors from NVIDIA and KAUST
SWAG members Nabil Masmoudi (PhD student) and Vladimir Kazei (Postdoctoral fellow) working on SOFI3D, a community seismic imaging code, were the most successful using OpenACC with managed memory to get the highest performance upgrade. Speed-up achieved (in comparison to the best performance on a dual sockets Haswell CPU server) was 1.15X with a Tesla K80 and 2.3X on Pascal-based Tesla P100 GPU.