1 reply to this topic

[bizertino]

Hi,

i m trying to implement a relatively complex algorithm with floating format data type, the hw module generated is bigger than the virtex II xc2vp30.
Now i m trying to change it to fixed point format and pricisely into costum fixed point 18bits width to correspond to the FPGA mutipliers. I m asking if i want to do some calculations with that width and format, i have to develop macros like in co_math to add multiply .. such data? And if the answer is yes, the hardware result will be economic or not ?

thx.

[RalphBodenner]

Yes, I would recommend using custom macros for your 18-bit type. You can use co_uint18 to contain the bits. Fixed-point arithmetic consumes much fewer resources than floating-point.

Ralph