This post is older than 2 years and might not be relevant anymore
More Info: Consider searching for newer posts

Tips for CMSIS-DSP simulation?

Hi All,

This may not be 100% nRF-related, but:

I have a set of algorithms that are written with Keil uVision 5 for nRF52840 using the CMSIS-DSP library. Now, I would need to simulate the exact output of these algorithms using an x86 computer. The optimal situation would be, that I could make test runs where I could route data from an input file to the algorithms, and write the output to another file. 

One solution could be to try to modify the CMSIS-DSP library so that it would compile for x86. Then the whole project could be made to compile on x86. I have done something like this in the history, but it is a lot of work, and not as good a proof for a working algorithm than running a real processor simulator with a real ARM binary.

Any tips for the easiest way forward? I think the need is quite common?

Parents
  • Hi,

    You can use the Jumper Virtual Lab (https://jumper.io) for simulating the ARM CMSIS-DSP library. Jumper Virtual Lab is an alternative for QEMU. It's an emulation platform that is built around a behavioral simulation model of the HW - the MCU core, MCU peripheral and external sensors, and peripherals. Jumper has a support for the nRF52 and the ARM CMSIS-DSP library.

    Jumper Virtual Lab allows you to run unit testing and full system tests with a Python SDK. It also supports rapid development and debugging capabilities (with integration to any GDB client).

    Let me know if you have any questions.

    Good luck!

    Yaniv

Reply
  • Hi,

    You can use the Jumper Virtual Lab (https://jumper.io) for simulating the ARM CMSIS-DSP library. Jumper Virtual Lab is an alternative for QEMU. It's an emulation platform that is built around a behavioral simulation model of the HW - the MCU core, MCU peripheral and external sensors, and peripherals. Jumper has a support for the nRF52 and the ARM CMSIS-DSP library.

    Jumper Virtual Lab allows you to run unit testing and full system tests with a Python SDK. It also supports rapid development and debugging capabilities (with integration to any GDB client).

    Let me know if you have any questions.

    Good luck!

    Yaniv

Children
No Data
Related