• State
  • Locked Locked
  • Replies 1 reply
  • Subscribers 98 subscribers
  • Views 348 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Efficient Matrix Operations in IQMath? (DSPLIB?)

Travis Wells
Prodigy 10 points

Hi all,

In short my question is: Is there a way to use to perform efficient matrix operations using IQMath to implement a floating point algorithm?

e.g. is it possible in someway to get DSPLIB to handle IQMath on the C6472?

Some background:

  1. I have a floating point algorithm (simulated/developed in MATLAB) and wish to implement it on the C6472 processor in fixed point.
  2. I am mainly working with Matrices so need matrix multiply/transpose operations. 
  3. Using the IQMath library seems the best way (in terms of development time/simplicity) to transfer from a floating point to fixed point implementation
  4. However there are no DSP/matrix functions in IQMath
  5. DSPLIB obviously is another way to go, but DSPLIB functions only handle 16 bit numbers.

I've written my own routines to do matrix operations in IQMath but obviously these are not as efficient as the DSPLIB routines. Have I overlooked something?

Appreciate any help/guidance to lead me in the right direction.

Thanks,
Travis

  • 0
    TI__Guru* 84110 points

    Travis,

    From the TMS320C64x DSP Library Programmer’s Reference, SPRU565, Appendix A.2:

    SPRU565B said:
    Unless specifically noted, DSPLIB functions use Q15 format, or to be more exact, Q0.15.

    Regards,
    RandyP

     

    If you need more help, please reply back. If this answers the question, please click  Verify Answer  , below.