[Libre-soc-dev] DCT/FFT augmentations

[Luke Kenneth Casson Leighton]

On 6/27/21, Richard Wilbur <richard.wilbur at gmail.com> wrote:

> a floating point number and then did the floating point multiply.  My
> recollection is that floating point addition requires more setup and
> renormalization than multiply.


fmac you need 3x the precision on the mantissa: the mul creates a
mantissa 2x its inputs.

we'll have to have this.

l.