Can GPGPU Programming Be Liberated from the Data-Parallel Bottleneck?
Advanced Micro Devices
IEEE Computer, vol. 45, no. 8, pages 42-52, 2012
@article{gaster2012can,
title={Can GPGPU Programming Be Liberated from the Data-Parallel Bottleneck?},
author={Gaster, B.R. and Howes, L.},
journal={Computer},
pages={42–52},
year={2012},
publisher={IEEE Computer Society}
}
Heterogeneous parallel primitives (HPP) addresses two major shortcomings in current GPGPU programming models: it supports full composability by defining abstractions and increases flexibility in execution by introducing braided parallelism.
September 7, 2012 by hgpu
