Efficient 2D FFT in C for Large Arrays: Where to Find the Code?

  • Thread starter Thread starter Maciej Olszewski
  • Start date Start date
  • Tags Tags
    2d Fft
Click For Summary
SUMMARY

The discussion centers on finding efficient C code for performing a 2D Fast Fourier Transform (FFT) on large arrays, specifically 8192 x 8192 dimensions. Users reported unsatisfactory results from the Paul Bourke and Sanfoundry resources, citing slow computation times. Recommendations include using FFTW, a widely recognized library for FFT operations, which also supports GPU implementations for enhanced performance. Additionally, the Ooura FFT library is suggested as an alternative for those seeking different options.

PREREQUISITES
  • Understanding of 2D Fast Fourier Transform (FFT) algorithms
  • Familiarity with C programming language
  • Knowledge of FFTW library and its functionalities
  • Basic concepts of GPU computing for FFT acceleration
NEXT STEPS
  • Explore the FFTW library documentation for implementation details
  • Investigate the Ooura FFT library for alternative FFT solutions
  • Learn about optimizing FFT performance on GPUs
  • Research best practices for handling large data arrays in C
USEFUL FOR

Software developers, researchers in signal processing, and anyone involved in computational mathematics seeking efficient methods for performing 2D FFTs on large datasets.

Maciej Olszewski
Messages
1
Reaction score
0
I am looking for a 2D fft that takes in a 2D array of heights and does a fft in c on the array. I would like if the code was capable of working with big arrays, such as 8192 x 8192 I have tried the paulbourke and sanfoundry websites. The first was not giving me the output that is expected and the second one took too long to compute. Does anyone know where I could find such code?
 
Technology news on Phys.org
On linux you have the libfft. I never use it but i know it exist.
 
As Dr Claude said FFTW, it is pretty well the standard, it even has GPU implementations if you have a decent graphics card.
 

Thread 'Use of AI (ML/DL) in Science'

We have many threads on AI, which are mostly AI/LLM, e.g,. ChatGPT, Claude, etc. It is important to draw a distinction between AI/LLM and AI/ML/DL, where ML - Machine Learning and DL = Deep Learning. AI is a broad technology; the AI/ML/DL is being developed to handle large data sets, and even seemingly disparate datasets to rapidly evaluated the data and determine the quantitative relationships in order to understand what those relationships (about the variaboles) mean. At the Harvard &...
View full post »

Similar threads

MATLAB [~,~,~,~,~,~,~,~,spec] Too many output arguments...
  • · Replies 5 ·
Replies
5
Views
2K
Python Shifting an array one postion to the left -- Non Function Vs Function
  • · Replies 17 ·
Replies
17
Views
3K
(CFD) Problem with C code for 1D linear convection equation
  • · Replies 1 ·
Replies
1
Views
2K
C# Missing samples recovery with FFT in C#
  • · Replies 2 ·
Replies
2
Views
3K
Populating a 1001x101 2D array in MATLAB
  • · Replies 3 ·
Replies
3
Views
2K
Python Getting this Array to be in 2D instead of 1D for Python Linear Regression
  • · Replies 22 ·
Replies
22
Views
7K
Why is My C Code Unexpectedly Changing Other Elements in a 2D Array?
  • · Replies 1 ·
Replies
1
Views
2K
Fortran Interpolating & Integrating Data from File to 2D Array
  • · Replies 10 ·
Replies
10
Views
26K
Fast reciprocal square root algorithm
  • · Replies 30 ·
2
Replies
30
Views
6K
Fortran [FORTRAN] FFT of delta function, issue w/ MKL & Intel compiler
  • · Replies 8 ·
Replies
8
Views
2K