----- Original Message -----
> Date: Mon, 07 Mar 2011 19:41:43 -0800
> From: "Chuck Guzis" <cclist at sydex.com>
> Subject: Re: hardware multiply/divide functionality in CPUs
>
> The 701 had multiply and divide instructions as did the 650. But
> then, the IBM 603 could multiply and the 604 could multiply and
> divide.
>
> --Chuck
----- Reply: -----
I suppose within the paradigm of the day you could consider the 603 and 604
"numeric co-processors", the only electronic (tubes of course) components in
an otherwise electro-mechanical "system".
The 603's claim to fame is that it's considered to be "the first
mass-produced commercial electronic calculating device" and the 604 which
came two years later in 1948 was an improved version which, as Chuck points
out, added division capabilities.
Having myself worked with the 604 (640 kilos/4 m3) I had to chuckle at its
being called 'a miniature electronic calculator'; the word 'miniature'
certainly had a different meaning in the context of the day:
http://www.science.uva.nl/faculteit/museum/604.php
Wish I'd kept my IBM manuals, including the 604; the person to whom I sent
them years ago (you know who you are ;-) solemnly promised to make them
available on the 'web, but I haven't found them so far...
BTW, FWIW with a little fancy programming the 40x series of 'tabulating
machines' could also be made to multiply, albeit *very* slowly... ;-)
mike
----- Original Message -----
>
> Date: Mon, 07 Mar 2011 15:27:46 -0800
> From: "Chuck Guzis" <cclist at sydex.com>
> Subject: Re: hardware multiply/divide functionality in CPUs
>
> There were math co-processor boards (AM9511, AM9512, for example) for
> S100 systems. ISTR that there was even a TRW bipolar (16x16? Huge
> chip that ran hot as a pistol) for the S100 bus.
>
> --Chuck
----- Reply: -----
Also Cromemco's Maximizer co-processor and the XXU 68020/68881 (and later
68040) processor boards.
Konrad Zuse's Z3 relay computer, built in 1941, had hard wired 22-bit floating point multiply, divide, and square root instructions. That is likely the first implementation of hardware multiply and divide in a computer. Multiply took up to 3 seconds.
The IBM 1130 had 16x16 multiply and divide in 1967. Shift and add.
It was not a microcoded machine. On average, a multiply took 26 usec,
and 11 of those were the overhead of fetching the instruction and
operands.
----- Original Message -----
>
> Date: Mon, 7 Mar 2011 15:02:57 -0800 (PST)
> From: Fred Cisin <cisin at xenosoft.com>
> Subject: Re: I Never Knew That The VT100 Was A Computer :-)
> It had some damage; I had to retrieve it after the college dumpstered MY
> collection (in spite of specific paperwork from the dean authorizing it!)
> I'd like to drop it on a college administrator's head.
> --
> Grumpy Ol' Fred cisin at xenosoft.com
----- Reply: -----
No wonder you're grumpy...
m
On 03/07/11 23:57, allison<ajp166 at verizon.net> wrote:
> EAE was PDP-8
Actually, Tony is totally correct. There was an EAE for the PDP-11 as
well, and it is a Unibus device, with a few registers in floating space.
You load some of them with the arguments, and then read out the result
at another register a little later.
It was produced before the EIS came about, and probably only sold with
the 11/20 and 11/15. But it should work on any Unibus machine.
> The PDP11 series had multiple implementations including the FPP (2901 based)
> and the FIS for the 11/23cpu (F11). The 11/44 had a FPU that used a carload
> of 2901s.
FIS first came on the 11/40. The same opcode space was then reused for
the FPP on the 11/45 and onwards.
11/23 never had FIS, but the 11/03 did. 11/23 had an optional FPP chip,
however (and CIS).
In fact, the only two machines to have FIS was the 11/40 and 11/03 (and
I know the 11/35 as well, but that is just a derivative of the 11/40).
Why the 11/03 got FIS is beyond me, but maybe they didn't have enough
transistors to implement the FPP, but wanted something, and reused that
design instead, since you at least have some software that supports it.
I'm pretty sure the FIS for 11/03 was an option, though, and not
included in every machine.
The FIS for the 11/40 definitely was.
As far as integer multiply/divide goes, the 11/40 had EIS as well, but
it was optional. Just like FIS.
Johnny
Eric writes:
> Issue #3 is unfortunately the final issue.
> You can now get PDF files of issues 1 through 3 from their web site:
> http://www.300baudmagazine.com/
And I never knew it existed until now :-(.
The enthusiastic attitude and WIDE pop-culture-artifact orientation
reminds me a lot of CHAC newsletters (did they/Kip Crosby do anything
past the mid-90's?)
Tim.
Yes, a whole stack of 5-1/4 inch floppy disks for your Victor computer.
All copies, no originals.
Are they any good? Who knows!
Do you want them? Heck yeah!
Pay shipping and we can make it so.
No sniping allowed.
