floating point conversion VHDL 2008

Question

I am trying to use the VHDL 2008 floating point libraries. When using the type conversion "to_float" is this converting to single precision or double precision?

Also, if there are any useful links or resources that talk about the floating point libraries in VHDL 2008 that would be a great help.

Answer 1

The second and third arguments of the three-argument form of to_float() are the exponent and fraction sizes. By passing appropriate values it is possible to generate either. The defaults are 8 and 23 respectively, which results in a float32.