# How to allocate DMA buffer for 2D FFT parallelization?
**URL:** https://discuss.pynq.io/t/how-to-allocate-dma-buffer-for-2d-fft-parallelization/5589
**Category:** Support
**Created:** [May 16, 2023, 3:57pm UTC](https://discuss.pynq.io/t/how-to-allocate-dma-buffer-for-2d-fft-parallelization/5589 "2023-05-16T15:57:26Z")
**Posts on this page:** 4
**Page:** 1
### Author:  [@Asmami](https://discuss.pynq.io/u/Asmami)
#### Post date: [May 16, 2023, 3:57pm UTC](https://discuss.pynq.io/t/how-to-allocate-dma-buffer-for-2d-fft-parallelization/5589/1 "2023-05-16T15:57:26Z")
Hi, I am implementing 2D FFT on my ZCU104 board and using PYNQ for testing. The image is unit8 512 by 512 . Since it is uint8, I am able to parallelize my design by streaming 256-bit at a time. The hardware design works, however, I have to reshape the image before copying it to the DMA buffer. This is because it reads a whole column before going to the next column making the rows change constantly. I want to have the same rows being read until all columns in the rows have been read.
Right now I am slicing and stacking my image from (512, 512) to (512\*16, 32) for it to work as intended. This task is “meh” and adds to pre-processing. Is there a way to tell the allocate function how the image is to be represented in memory without the need for slicing and stacking? I have looked around the forums, but I have not found an answer.
Best regards,
Asbjørn
---
### Author:  [@marioruiz](https://discuss.pynq.io/u/marioruiz)
#### Post date: [May 17, 2023, 8:30am UTC](https://discuss.pynq.io/t/how-to-allocate-dma-buffer-for-2d-fft-parallelization/5589/2 "2023-05-17T08:30:34Z")
Hi @Asmami,
Welcome to the PYNQ community.
> Is there a way to tell the allocate function how the image is to be represented in memory without the need for slicing and stacking?
No, `order` is not supported by the `allocate`, I believe this is a constraint from the driver. But, you could do something like this using `numpy`
```python
shape = (2,2)
buf0 = allocate(shape=shape, dtype=np.int8)
buf1 = allocate(shape=shape, dtype=np.int8)
for i in range(shape[0]):
for j in range(shape[1]):
buf0[i][j] = i * shape[0] + j
# Transpose (change rows and columns order)
buf1[:] = buf0.T
print(f'{buf0=}\n{buf1=}')
```
Or you can try with reshaping the array, I suggest you use the built-in numpy APIs which are super efficient.
Mario
---
### Author:  [@Asmami](https://discuss.pynq.io/u/Asmami)
#### Post date: [May 18, 2023, 3:32pm UTC](https://discuss.pynq.io/t/how-to-allocate-dma-buffer-for-2d-fft-parallelization/5589/3 "2023-05-18T15:32:36Z")
Thanks for the quick answer. It was not what I hoped for, however, it was expected. Thanks for the optimization tips.
Asbjørn
---
### Author:  [@system](https://discuss.pynq.io/u/system)
#### Post date: [May 21, 2023, 3:33pm UTC](https://discuss.pynq.io/t/how-to-allocate-dma-buffer-for-2d-fft-parallelization/5589/4 "2023-05-21T15:33:36Z")
This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.