• State
  • Locked Locked
  • Replies 4 replies
  • Subscribers 79 subscribers
  • Views 969 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.

How to use Hardware Multiplier for Floating Point calculations?

Mrt_Nsk
Intellectual 270 points
Other Parts Discussed in Thread: MSP430F2417

Hello everyone,

I am using a MSP430F2417 device and CCE. I am working on a code with extensive floating point calculations. The floating point arithmetic calculation routines in the run time support library do not use the hardware multiplier present on the chip.

Does anyone know how to use the hardware multiplier for the floating point calculations in CCE?

Thanks!

  • 0
    TI__Mastermind 24390 points

    Hi,

    You may want to refer to the following app note, which shows how to efficiently multiply using general MSP430 multiplication. Floating point is addressed: http://focus.ti.com/mcu/docs/mcusupporttechdocsc.tsp?sectionId=96&tabId=1502&abstractName=slaa329

    Edit: I am sorry, this report is only for software multiplication! The only examples using the MPY I know of are on the main code site www.ti.com/msp430codeexamples

    Sorry for the confusion!

  • 0 in reply to Brandon08
    Intellectual 270 points

    Hi Brandon,

    Thanks for the post.

    The MSP430 code examples do not have a single code example showing the use of the hardware multiplier for floating point calculations.

    Any other link of examples depicting the use of hardware multiplier for the floating point calculations? I am running against a deadline and i need to solve this issue urgently!

    Please help!

    Thanks.

  • 0 in reply to Brandon08
    Intellectual 270 points

    hello everyone,

    If we cannot use the hardware multiplier for the floating point calculations, does anyone know of any other method by which we can improve the execution speed of the floating point calculations? Can we write any other multiplication routine using C for efficient FP multiplication?

    I have a lot of floating point calculations in my code and the values are also big, thus it takes a lot of time for the execution of the code. How should i improve this?

    Thanks!

  • 0 in reply to Mrt_Nsk
    Prodigy 10 points

    Considered writing your own C library with inline assembly that uses the hardware multiplier? surely it would be 50 lines of code max to implement something like this. Find out how to do inline assembly first, then look up the assembly instructions for the operation you require.

    edit: whoops this was a 2008 post not 2009

**Attention** This is a public forum