cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
sha@hys
Explorer
Explorer
3,654 Views
Registered: ‎05-31-2015

passing float values from microblaze to VHDL

Jump to solution

Hello,

 

   I am doing a PID loop in VHDL. I am passing values from computer to microblaze and then microblaze to VHDL. I pass float to microblaze. Can I pass float values without any conversion to VHDL? My process is time critical .When I go for float to VHDL required format conversion my time criticality is not met. Or else any way to speed up float calculations in microblaze. My microblaze use floating point units as well as cache units. Any suggestions welcome.

 

Thanks in advance

 

With regards

Shalini

0 Kudos
1 Solution

Accepted Solutions
sha@hys
Explorer
Explorer
5,765 Views
Registered: ‎05-31-2015

Hello,

 

Finally I gave floating point representation of number to the floating point core in VHDL, that convert it to required fixed point format.

 

 

View solution in original post

0 Kudos
3 Replies
u4223374
Advisor
Advisor
3,629 Views
Registered: ‎04-26-2015

Floating-point calculations can't be automatically synthesized in VHDL (or Verilog). You can certainly pass a floating-point value out to the VHDL code (where it'll just be treated like a 32-bit integer) and then pass that into the Xilinx Floating Point IP core for further calculations - but this seems like it's potentially more trouble than it's worth.

0 Kudos
sha@hys
Explorer
Explorer
3,617 Views
Registered: ‎05-31-2015

Hello,

 

Thank you for reply. 

 

"You can certainly pass a floating-point value out to the VHDL code (where it'll just be treated like a 32-bit integer)"

 

What is the integer I should pass? IEEE754 representation of floating point number I want. Can this integer be given to floating point IP core?

 

With Regards

shalini

0 Kudos
sha@hys
Explorer
Explorer
5,766 Views
Registered: ‎05-31-2015

Hello,

 

Finally I gave floating point representation of number to the floating point core in VHDL, that convert it to required fixed point format.

 

 

View solution in original post

0 Kudos