Forums - FastCV for Linaro

6 posts / 0 new
or Register
Last post
FastCV for Linaro
ilyak
Join Date: 28 Dec 14
Posts: 3
Posted: Mon, 2015-02-16 03:58
Top

Hi 

We are planning to develop Image processing system on Linaro IFC6410 (Snapdragon 600).

Can we use FastCV (accelerated for Snapdragon) in Linaro environment?

What kind of acceleration we have there - Hexagon or Adreno? 

Thank you

I found the answer in FastCV release notes - 

FastCV 1.2.2 has the capability to select the QDSP or GPU implementation when using
QDSP and GPU supported Qualcomm device.
 
FastCV 1.2.2 will automatic select the best implementation for each API based on the
operation mode chosen. 
 
Operation mode                          Description
 
FASTCV_OP_LOW_POWER       The QDSP implementation will be used unless the
                                                     QDSP speed is 3 times slower than CPU speed.
FASTCV_OP_PERFORMANCE   The fastest implementation will be used.
FASTCV_OP_CPU_OFFLOAD    The QDSP implementation will be used when it’s
                                                     available, otherwise it will find for GPU and CPU  implementation.
 
 
 
  • Up0
  • Down0
Re: FastCV for Linaro
jeff4s Moderator
Join Date: 4 Nov 12
Posts: 106
Posted: Wed, 2015-02-18 14:54
Top

Hi,

FastCV doesn't support Linaro. There are Android and PC versions of the library only.

Cheers,

-Jeff

  • Up0
  • Down0
Re: FastCV for Linaro
ilyak
Join Date: 28 Dec 14
Posts: 3
Posted: Sat, 2015-02-21 23:20
Top

 Hi, Jeff

 Thank you for the answer.

 Can you give me a simple way to get started with FastCV

 on IFC6410?

 At the moment our goal is to estimate convolution time for APQ8064.

 using FastCV (or OpenCL)

Regards,

Ilya

 

 

  • Up0
  • Down0
Re: FastCV for Linaro
ilyak
Join Date: 28 Dec 14
Posts: 3
Posted: Sun, 2015-02-22 00:12
Top

Sorry about my persistence.

We need evaluation only. Can we use this - 

------------------------------------------------------------------------

FastCV SDK (v1.2.2) for Linux OS
Download (9.7 mb)
Updated 03 Mar 14
This release is an evaluation version of FastCV 1.2.2 that is built to execute on a target which implements the Linux Operating system. This release is intended to be used for evaluation purposes with the understanding that on-going releases of FastCV are not currently planned.
 
------------------------------------------------------------------------
from https://developer.qualcomm.com/mobile-development/add-advanced-features/...
 
 Thanks,
ilya
  • Up0
  • Down0
Re: FastCV for Linaro
jeff4s Moderator
Join Date: 4 Nov 12
Posts: 106
Posted: Sun, 2015-02-22 10:06
Top

Hi,

Can you use Android? If yes the Dragon Board is a good option for evaluation.

FastCV 1.2.2 may or may not have the APIs depending on what you want to evaluate.

Cheers,

-Jeff

  • Up0
  • Down0
Re: FastCV for Linaro
lucasdsfranco
Join Date: 9 May 16
Posts: 1
Posted: Tue, 2016-08-02 07:30
Top

Hi,

I'm having the same problem. I have one IFC6540 board running a application for image processing under linaro, i want to optimize my code. What are my options?
- OpenCL is not supported on linaro

- Is FastCV supported on linaro?

What can I do to use heterogeneos computer power to optimize my image processing?

In my software I dont need any graphical services, it processes images only in textmode. I want to use heterogeos processing, opencl, fastcv or some other tool to optimize the time processing.

What snapdragon 805 supports on linux?

 

Thanks

Lucas

  • Up0
  • Down0
or Register

Opinions expressed in the content posted here are the personal opinions of the original authors, and do not necessarily reflect those of Qualcomm Incorporated or its subsidiaries (“Qualcomm”). The content is provided for informational purposes only and is not meant to be an endorsement or representation by Qualcomm or any other party. This site may also provide links or references to non-Qualcomm sites and resources. Qualcomm makes no representations, warranties, or other commitments whatsoever about any non-Qualcomm sites or third-party resources that may be referenced, accessible from, or linked to this site.