Prev: nonlinear kalman filter
Next: how to do Auto-corelation and crosscorelation
From: Calendini Pierre on 3 May 2010 17:48
Recent embedded processor are able to use 32b simple precision float calculation with integrated hardware FPU.
How force simulink model to use also simple precision computation to be close to real calculation done in such processor ? (I understand that native computation use 64b floating point).
Thanks
From: Steven Lord on 4 May 2010 10:19

"Calendini Pierre" <poc.calendini(a)free.fr> wrote in message
news:hrngak$n92$1(a)fred.mathworks.com...
> Recent embedded processor are able to use 32b simple precision float
> calculation with integrated hardware FPU. How force simulink model to use
> also simple precision computation to be close to real calculation done in
> such processor ? (I understand that native computation use 64b floating
> point).

I suspect from skimming the contents of the chapter that the answer is in
this chapter in the documentation:

http://www.mathworks.com/access/helpdesk/help/toolbox/simulink/ug/f14-90479.html

--
Steve Lord
slord(a)mathworks.com
comp.soft-sys.matlab (CSSM) FAQ: http://matlabwiki.mathworks.com/MATLAB_FAQ


 | 
Pages: 1
Prev: nonlinear kalman filter
Next: how to do Auto-corelation and crosscorelation