32 bit Floating Point for Micro Controllers

09 Aug

There are times in the life of embedded system designers and programmers when the absence of a floating point unit is very strongly felt. The processing power of these tiny, space-saving micro controllers, frequently becomes a limiting factor if some hefty mathematical computation is required. Although we have been seeing chips like Blackfin and dsPIC that combine µC features with powerful DSP processors but the kind of price involved is a hindrence in using them for smaller projects. it looks like someone came up with a superb solution to the problem.

um-FPU block diagram

um-FPU block diagram

MicroMegacorp have created an FPU that comes in two versions an 8-pin DIP and an 18-pin DIP package. They can be connected via an I²C or SPI interface with virtually any micro controller that supports either of these interfaces. An IDE is provided by Micromegacorp that allows convenient translation of mathematical expressions into uM FPU code. The uM FPU is also capable of performing fixed point,floating point,long integer addition, multiplication,division and type casting functions. Not just that, it comes with flash memory that can be used for user defined functions.

I’ll definitely recommend using this space saving number cruncher for your next project. Do share your experience in replies to this post.


Tags: , , , , , , , , , , , , , , , , , ,

Leave a Reply


  1. Faried Nawaz

    August 19, 2008 at 11:28 am

    MicroMegacorp? What a name!

  2. Abbas Ahmed

    August 22, 2008 at 10:14 pm

    🙂 They make microchips that perform mega tasks …