This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Floating Point capabilities of CC32xx

Other Parts Discussed in Thread: CC3200

Hi,

I would like to know how is the behaviour of a CC32xx with floating points management. I know it does not have a FPU however it is mentioned as a ARM Cortex M4F in many places.

The main requirements of my device would be:

1- Wireless communication using WiFi. The device will exchange logs and parameters with a browser using WebSockets. We can try with BLE too as we are aiming phone/tablets.

2- Sound generation and output to Headphones. I have to create a vector from scratch using formulas like "sin()", "rnd()", "log()", "sqrt". Sounds like Pure Tones and filtered white noises will have to be created and played in real time.

3- Get sensors data using i2c or SPI, controlling an audio codec using SPI, and flash memory access to store the logs.

4- Floating point operations is preferable as we want the best quality

5- Consume less power as it is possible. It will be battery powered.

I guess an ARM Cortex M4F would be ideal for me as a first approach. And having i2s and WiFi support as the CC32xx has is great but it doesn't support floats in hardware.

Are there any alternative that I could use instead of the CC32xx? I tried to find an ARM Cortex M4F with i2s in TI but didn't have enough luck.

  • Hi Leinho,
    CC3200 M4 do not support Floating Point Hardware. Can you please point me to TI documentation which mentions M4F instead, will get it fixed.
    Regards,Geet
  • No, it doesn't say so there. I wanted to ask if there were some work around or maybe through firmware but that for sure will be slower than in hardware. It says that outside, it wikipedia looking for ARM M processors.

    Do you recommend me any alternative to this device but with FPU?

    My main goal is:

    - Sound generation and output to Headphones. I have to create a vector from scratch using formulas like "sin()", "rnd()", "log()", "sqrt". Sounds like Pure Tones and filtered white noises. This will have to be sent to the earphones and the generation can be modified in real time (like changing the frequency of the pure tone and the output has to be modified accordingly).

    I looked into the TIVA series but they are not for audio and don't support i2s. I may have to follow the paper that "creates one" using SPIs but that may not be efficient enough.

  • Alternatively if operations are not very inensive, try using floating point libraries on CC3200.

    Regards,
    Geet