My primary reference apropos these is the debate between gustafson and kahan
(of ieee-754 fame): http://www.johngustafson.net/pdfs/DebateTranscription.pdf
I can follow the arguments, but have difficulty evaluating them. I don't
really know what to make of it.
On Tue, 18 Oct 2022, Jan-Pieter Jacobs wrote:
If my memory serves me well, I remember Gustafson's posit number format
being discussed here.
Today, I found an article in IEEE spectrum about some tests of a hardware
implementation of it, that might be of interest to some of you:
https://spectrum.ieee.org/floating-point-numbers-posits-processor
Best regards,
Jan-Pieter
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
