• State TI Thinks Resolved
  • Locked Locked
  • Replies 1 reply
  • Answers 1 answer
  • Subscribers 79 subscribers
  • Views 675 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

Original question:

IQ math data types

MSP430FR4133: QMath Library Advantage

Mesut TOPUZLU
Intellectual 690 points
Part Number: MSP430FR4133
Other Parts Discussed in Thread: MSP430FR5889

Hi.

My question is about QMath lib. In the Qmaths user guiede described that  

"The Qmath library takes advantage of the MPY32 multiplier peripheral when it is available. If the
device does not have the MPY32 peripheral then the CPU is used to perform a software multiply."

Now I am using MSP430FR4133 chip that has no hardware multipler. But, there is also MPYSoftware lib. for

chips that have no hardware multipler.

In this case, does using the MPYSoftware QMathLib provide an advantage for performance and power consumption.

Because in my project I am using floating point values and ULP Advisor warnings me constantly. :)

If there is no advantage for MPYSoftware, I will migrate to MSP430FR5889 etc. chip.

  • 0
    Guru 26200 points
    Hello Mesut,

    QMath and IQMath libraries are not floating point libs.
    Within this document: dev.ti.com/.../MSP432_IQmathLib-UsersGuide.pdf
    on pages starting from 7 you will find description of the Q and IQ formats used by these libraries.
    Starting from page #11 you will find that there are separate library files for MCUs having MPY32 and those not having MPY32.

    You have ULP Adviser warnings because FP operations are heavy crunching tasks for tiny 16 bits MSP430 MCUs.

    You wrote: If there is no advantage for MPYSoftware, I will migrate to MSP430FR5889 etc. chip.
    I say: No MPY32, no advantage.

    You do not provide any information on how intensive FP operations you need.
    One time per hour, per second?

    Do tests how fast is floating point on your MSP430FR4133.
    Do not expect big boost for floating point when using MPY32 because your code will not multiply all the time.
    No guarantee that transition to 5889 would move a step close to a solution you need.

    If you really need FP go to MSP432P401R or any other M4F based MCU.

**Attention** This is a public forum