17 Oct
2007
17 Oct
'07
6:01 p.m.
Subject: Re: FPUs for small computers (Was: What to download for a PDP-8)
From: "Chuck Guzis" <cclist at sydex.com>
Date: Wed, 17 Oct 2007 10:21:40 -0700
To: "General Discussion: On-Topic and Off-Topic Posts" <cctalk at
classiccmp.org>
On 17 Oct 2007 at 12:09, Jim Battle wrote:
If it's the article as I remember it it was a push for Add/subtract
but better for multiply and divide (to 8 digits). An 8080 does a
16x16 multiply in a few milliseconds to compare.
In another Byte there was a simple design and explanation of
hardware multiply by shift and add. That was capable of high speed.
With TTL of the time that was something under 10 microseconds
for 16 bits using a 2mhz clock. At that point a 4mhz Z80 takes
nearly that much time to load and read it. You can do subtract
and shift to implement a divide.
Allison
Finally, I recall seeing an article where
somebody took a pocket
calculator chip and essentially poked simulated keystrokes at it and
then decoded the LED driver output to determine the answer. It was very
slow, though, so all it saved was the space of the floating point
library code.
I remember that one. It may have been an early issue of Byte. I
also wondered at the time if any speed was gained over just running
plain old x80 code.