• State Resolved
  • Locked Locked
  • Replies 1 reply
  • Subscribers 98 subscribers
  • Views 247 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.

Optimized DOT product on L138

vikas sharma
Prodigy 230 points

We would like to use an optimized dot product library routine for C674x L138 (ELF format) and wanted to know which of the following we should use

a. C64XPLUS-DSPLIB: C64x+ DSP Library   ( DSP_dotprod ) Vector Dot Product

b. c64xplus-iqmath_2_01_04_00

On C55x we could use the intrinsics which provided fast and efficient computaion.

Looking for something similar for c674x  or if anyone can point me to assembly code for the same.

Regards,

-Vikas

  • 0
    TI__Guru** 116170 points

    Vikas,

    Both the libraries that you have mentioned provide you functions to perform vector dot products. The C64x DSPLIB will provide a fixed point vector dot product while the IQMATH will provide you functions by which you can perform vector dot product of vectors with floating point co-efficient with fixed point math. The IQMATH library is generally used only on fixed point C64x+ core on C674x which is a floating and fixed point processor, this is not required. If the vectors are floating point, I would recommend you to look at the C674x version of the DSPLIB for this functionality.

    http://www.ti.com/tool/sprc265

    If you would like to create your own function using intrinsics, please refer to the C6000 Compiler documentation that are provided in the docs folder of the C6000 compiler.

    Regards,

    Rahul