MSP430FR2353: Floating Point Operations

Jing Gloria

Part Number: MSP430FR2353

I have to code some formulas which are originally using Floating Point values. I understand that the FR2353 has no Floating Point hardware.

So far all I need are the four basic operators (+, -, *, /). Should I just code these as FP? Only a couple of formulas. Code block is in a loop that gets executed once a second.

Or should I do the math in integer (scaling the values like using mV instead if V)?

Is there any publication addressing when to use FP vs Integer operations?

Thanks,

Jing (Julian) Gloria

over 3 years ago

0 Keith Barkley over 3 years ago

Guru 42115 points

Your best bet is to try and see if there are performance bottlenecks. Don't get fancy unless you know you need too.

0 Luke Ledbetter over 3 years ago

TI__Mastermind 25700 points

Hi Julian Gloria,

It mostly depends on your needs and what trade offs you want. You can scale everything up so you are essentially doing integer math, you can use one of the Math Libraries, you could also separate the parts depending on how the math works. So separate the whole number and the decimal part of the floating point. Onto Keith's point, if you address the bottleneck you can know which to choose. Do you need speed? Low on Memory? neither and just want it to be easy to code? That will help you decide which method to go with.

Regards,

Luke

0 Joe G. over 3 years ago

TI__Mastermind 25966 points

Example around 2:00 of video might help:

https://training.ti.com/rapid-proto-adv-topic-layered-energia-code

**Attention** This is a public forum

MSP low-power microcontrollers

MSP low-power microcontroller forum

MSP430FR2353: Floating Point Operations