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.

C6727 to C6745 migration

 

Hi,

 

We are interested in migrate our c6727 codes to a c6745 DSPs. We use mainly floating point routines (adaptative filters, eco cancellers, clock recovery system, and so on).

We've seen in the data sheet that 6747 is compatible with c67x+ set of instructions and also that it has a large MFLOPS figure.

We'd like to know if the 6745 will be capable of improving our C6727 performance which is almost full in processing time between real time interrupts (90%) and IRAM (95%).

 

We guess that it will be even better (faster and more IRAM) but due to this fixed point compatibility we are wondering if this can cause a poorer floating performance compared to a pure floating processor such as c6727.

 

Any experience comparing this DSPs in intensive floating point processing???

 

Thanks a lot

  • Alex,

    The C6745 executes faster than the C6727 and has architectural improvements. I would recommend any user to feel confident that they ccould make this migration and get improved performance.

    The way to know for sure is to compare some routines on the CCS Device simulators.

    The C674x core is a superset of the C67x+ floating point core in the C6727 and of the C64x+ fixed point core. You can learn more about this by reading the C674x DSP CPU & Instruction Set User's Guide. Take a look at Table A-1, for example, which shows instruction compatibility.

    Regards,
    RandyP

     

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