Calculation of cross-correlation function accelerated by TensorFloat-32 Tensor Core operations on NVIDIA's Ampere and Hopper GPUs.
Kohei FujitaTakuma YamaguchiYuma KikuchiTsuyoshi IchimuraMuneo HoriLalith MaddegedaraPublished in: J. Comput. Sci. (2023)
Keyphrases
- correlation function
- graphics processing units
- fourier transform
- graphics hardware
- gpu implementation
- general purpose
- graphics processors
- quality assurance
- parallel implementation
- cpu implementation
- image segmentation
- parallel processing
- massively parallel
- post processing
- multi view
- multiresolution
- high quality
- three dimensional