UPGRADE YOUR BROWSER

We have detected your current browser version is not the latest one. Xilinx.com uses the latest web technologies to bring you the best online experience possible. Please upgrade to a Xilinx.com supported browser:Chrome, Firefox, Internet Explorer 11, Safari. Thank you!

cancel
Showing results for 
Search instead for 
Did you mean: 
827 Views
Registered: ‎03-15-2018

fixed vs float resource usage

Jump to solution

Hello,

 

Synthesized a (32 x 32) multiplier ip,

mul construction - using LUTs

optimization option - speed

pipeline stages - 6

 

Also synthesized a 32 x 32 floating point ip

with no dsp usage.

 

Design run displays:

 LUTFF
float 32626681
fixed 3211021143

 

Is this justified that fixed point is taking more 

resource than float operation.

 

Also, how do I use multiplier ip for fixed point case

and for integer case?

 

(Am a starter w.r.t Xilinx Vivado and FPGA architecture) 

 

Many thanks in advance,

Varsha

0 Kudos
1 Solution

Accepted Solutions
Scholar jmcclusk
Scholar
1,163 Views
Registered: ‎02-24-2014

Re: fixed vs float resource usage

Jump to solution

Floating point actually only has a 24 bit mantissa, so you only get a 24x24 multiplier with floating point,  while 32 bit integer data generates a true 32x32 multiplier.   So these LUT usage results are quite possible.

Don't forget to close a thread when possible by accepting a post as a solution.
4 Replies
Scholar jmcclusk
Scholar
1,164 Views
Registered: ‎02-24-2014

Re: fixed vs float resource usage

Jump to solution

Floating point actually only has a 24 bit mantissa, so you only get a 24x24 multiplier with floating point,  while 32 bit integer data generates a true 32x32 multiplier.   So these LUT usage results are quite possible.

Don't forget to close a thread when possible by accepting a post as a solution.
769 Views
Registered: ‎03-15-2018

Re: fixed vs float resource usage

Jump to solution

Thank you for the reply, it helped a lot :)

 

I have another question.

How do I customize the multiplier ip

between fixed and integer type?

 

Is it that customizing the ip is the same for

fixed or integer type , but we interpret the

decimal point location?

 

Many thanks in advance,

Varsha 

0 Kudos
Scholar richardhead
Scholar
751 Views
Registered: ‎08-01-2012

Re: fixed vs float resource usage

Jump to solution

@varshajhapillai

Fixed arithmatic is really just integer arithmetic. The separation between integer and fractional part is really just a fixed offset in the value (ie. shift left by N bits before calculation then shift right by N after calculation). This shift is basically free in FPGAs.

 

 

744 Views
Registered: ‎03-15-2018

Re: fixed vs float resource usage

Jump to solution

Thank you so much, that makes it clear :)

0 Kudos