Optimizing the Fast Fourier Transform Using Mixed Precision on Tensor Core Hardware

Sorna, Anumeena; Cheng, Xiaohe; D'azevedo, Ed; Wong, Kwai; Tomov, Stanimire

doi:10.1109/HiPCW.2018.8634417

Title: Optimizing the Fast Fourier Transform Using Mixed Precision on Tensor Core Hardware

Conference · Sat Dec 01 00:00:00 EST 2018

DOI:https://doi.org/10.1109/HiPCW.2018.8634417· OSTI ID:1559731

Sorna, Anumeena ^[1]; Cheng, Xiaohe ^[2];

^[3]; Wong, Kwai ^[3]; Tomov, Stanimire ^[4]

National Institute of Technology Tiruchirappalli
The Hong Kong University of Science and Technology
ORNL
University of Tennessee, Knoxville (UTK)

The Fast Fourier Transform is a fundamental tool in scientific and technical computation. The highly parallelizable nature of the algorithm makes it a suitable candidate for GPU acceleration. This paper focuses on exploiting the speedup due to using the half precision multiplication capability of the latest GPUs' tensor core hardware without significantly degrading the precision of the Fourier Transform result. We develop an algorithm that dynamically splits the input single precision dataset into two half precision sets at the lowest level, uses half precision multiplication, and recombines the result at a later step. This work paves the way for using tensor cores for high precision inputs.

View Conference

Cite

Export

Save

Research Organization:: Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)

Sponsoring Organization:: USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR)

DOE Contract Number:: AC05-00OR22725

OSTI ID:: 1559731

Resource Relation:: Conference: Paralle Fast Fourier Transforms (PFFT18) - Bengaluru, , India - 12/17/2018 5:00:00 AM-12/20/2018 5:00:00 AM

Country of Publication:: United States

Language:: English

References (6)

Fast Fourier Transforms Gentleman, W. M.; Sande, G. Proceedings of the November 7-10, 1966, fall joint computer conference on XX - AFIPS '66 (Fall) https://doi.org/10.1145/1464291.1464352	conference	January 1966
Mixed Precision Iterative Refinement Techniques for the Solution of Dense Linear Systems Buttari, Alfredo; Dongarra, Jack; Langou, Julie The International Journal of High Performance Computing Applications, Vol. 21, Issue 4 https://doi.org/10.1177/1094342007084026	journal	November 2007
SPFP: Speed without compromise—A mixed precision model for GPU accelerated molecular dynamics simulations Le Grand, Scott; Götz, Andreas W.; Walker, Ross C. Computer Physics Communications, Vol. 184, Issue 2 https://doi.org/10.1016/j.cpc.2012.09.022	journal	February 2013
Accelerating scientific computations with mixed precision algorithms Baboulin, Marc; Buttari, Alfredo; Dongarra, Jack Computer Physics Communications, Vol. 180, Issue 12 https://doi.org/10.1016/j.cpc.2008.11.005	journal	December 2009
Fourier Transform Infrared Spectroscopic Analysis of Protein Secondary Structures Kong, Jilie; Yu, Shaoning Acta Biochimica et Biophysica Sinica, Vol. 39, Issue 8 https://doi.org/10.1111/j.1745-7270.2007.00320.x	journal	August 2007
Implementation of a fast Fourier transform (FFT) for image processing applications Buijs, H.; Pomerleau, A.; Fournier, M. IEEE Transactions on Acoustics, Speech, and Signal Processing, Vol. 22, Issue 6 https://doi.org/10.1109/TASSP.1974.1162620	journal	December 1974

Similar Records

Mixed-precision iterative refinement using tensor cores on GPUs to accelerate solution of linear systems

Journal Article · Wed Nov 25 00:00:00 EST 2020 · Proceedings of the Royal Society. A. Mathematical, Physical and Engineering Sciences · OSTI ID:1559731

Haidar, Azzam; Bayraktar, Harun; Tomov, Stanimire; +2 more

APNN-TC: Accelerating Arbitrary Precision Neural Networks on Ampere GPU Tensor Cores

Conference · Sun Nov 14 00:00:00 EST 2021 · OSTI ID:1559731

Feng, Boyuan; Wang, Yuke; Geng, Tong; +2 more

MULTI-CORE AND OPTICAL PROCESSOR RELATED APPLICATIONS RESEARCH AT OAK RIDGE NATIONAL LABORATORY

Conference · Tue Jan 01 00:00:00 EST 2008 · OSTI ID:1559731

Barhen, Jacob; Kerekes, Ryan A; ST Charles, Jesse Lee; +1 more

Title: Optimizing the Fast Fourier Transform Using Mixed Precision on Tensor Core Hardware

Citation Formats

References (6)

Similar Records

Related Subjects