NOTICE: Due to a lapse in annual appropriations, most of this website is not being updated. Learn more.
Form submissions will still be accepted but will not receive responses at this time. Sections of this site for programs using non-appropriated funds (such as NVLAP) or those that are excepted from the shutdown (such as CHIPS and NVD) will continue to be updated.
An official website of the United States government
Here’s how you know
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
Secure .gov websites use HTTPS
A lock (
) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.
Parallel MPI image reconstructions in GPU using CUDA
Published
Author(s)
Klaus Natorf Quelhas, Mark-Alexander Henn, Ricardo Farias, Weston L. Tew, Solomon I. Woods
Abstract
This work shows that it is possible to obtain faster MPI image reconstructions by implementing the algorithms in parallel in Graphics Processing Units (GPUs) using NVIDIA's CUDA (Compute Unified Device Architecture). While the parallel Kaczmarz's algorithm was slower than its serial version running in the Central Processing Unit (CPU), the parallel version of the Conjugate Gradient Normal Residual (CGNR) algorithm was about 58 times faster than its serial implementation, and about 10 times faster than the serial implementation of Kaczmarz's.
Citation
International Journal on Magnetic Particle Imaging
Natorf Quelhas, K.
, Henn, M.
, Farias, R.
, Tew, W.
and Woods, S.
(2023),
Parallel MPI image reconstructions in GPU using CUDA, International Journal on Magnetic Particle Imaging, [online], https://doi.org/10.18416/IJMPI.2023.2303043, https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=936159
(Accessed October 8, 2025)