Saved in:
Bibliographic Details
Main Authors: Deveshwar, Nikhil, Rajagopal, Abhejit, Larson, Peder E. Z.
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2512.04317
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915802400161792
author Deveshwar, Nikhil
Rajagopal, Abhejit
Larson, Peder E. Z.
author_facet Deveshwar, Nikhil
Rajagopal, Abhejit
Larson, Peder E. Z.
contents A mixed precision Fast Fourier transform (FFT) implementation is presented. The procedure uses per-block microscaling (MX), a global power-of-two prescale, and prequantized low bit twiddles. We evaluate forward and round-trip FFT fidelity on two public MRI datasets and compare the effect of various low precision formats, image sizes, and MX block sizes on image quality. Results show that mantissa precision is the primary limiter under MX scaling while ablations suggest weak dependence on image size but a clear block-size trade-off with larger block sizes resulting in better numerical performance.
format Preprint
id arxiv_https___arxiv_org_abs_2512_04317
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Mixed Precision FFT with applications in MRI
Deveshwar, Nikhil
Rajagopal, Abhejit
Larson, Peder E. Z.
Signal Processing
A mixed precision Fast Fourier transform (FFT) implementation is presented. The procedure uses per-block microscaling (MX), a global power-of-two prescale, and prequantized low bit twiddles. We evaluate forward and round-trip FFT fidelity on two public MRI datasets and compare the effect of various low precision formats, image sizes, and MX block sizes on image quality. Results show that mantissa precision is the primary limiter under MX scaling while ablations suggest weak dependence on image size but a clear block-size trade-off with larger block sizes resulting in better numerical performance.
title A Mixed Precision FFT with applications in MRI
topic Signal Processing
url https://arxiv.org/abs/2512.04317