Skip to Main content Skip to Navigation
Conference papers

On the Use of Small 2D Convolutions on GPUs

Abstract : Computing many small 2D convolutions using FFTs is a basis for a large number of applications in many domains in science and engineering, among them electromagnetic diffraction modeling in physics. The GPU architecture seems to be a suitable architecture to accelerate these convolutions, but reaching high application performance requires substantial development time and non-portable optimizations. In this work, we present the techniques, performance results and considerations to accelerate small 2D convolutions using CUDA, and compare performance to a multi-threaded CPU implementation. To improve programmability and performance of applications that make heavy use of small convolutions, we argue that two improvements to software and hardware are needed: FFT libraries must be extended with a single convolution function and communication bandwidth between CPU and GPU needs to be drastically improved.
Document type :
Conference papers
Complete list of metadata

Cited literature [3 references]  Display  Hide  Download
Contributor : Ist Rennes Connect in order to contact the contributor
Submitted on : Monday, June 21, 2010 - 3:10:44 PM
Last modification on : Saturday, November 18, 2017 - 6:16:02 PM
Long-term archiving on: : Wednesday, September 22, 2010 - 6:08:40 PM


Files produced by the author(s)


  • HAL Id : inria-00493873, version 1



Shams A.H. Al Umairy, Alexander S. van Amesfoort, Irwan D. Setija, Martijn C. van Beurden, Henk J. Sips. On the Use of Small 2D Convolutions on GPUs. A4MMC 2010 - 1st Workshop on Applications for Multi and Many Core Processors, Jun 2010, Saint Malo, France. ⟨inria-00493873⟩



Record views


Files downloads