Searching \ for 'Oldie question: IBM-709?' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=ibm+709
Search entire site for: 'Oldie question: IBM-709?'.

Truncated match.
PICList Thread
'Oldie question: IBM-709?'
1998\08\31@203449 by Mel Evans

picon face
   I recently heard someone say: "The PIC 16C54 has the same performance,
speed- and memory-wise, as the IBM-709".  The UNIVAC, maybe, but the IBM-709?
   OK, you oldtimers, here's a more general question:
       "The PIC 16C54 has the same performance, speed- and memory-wise, as
the IBM-x or the PDP-y."
   What are the values of x and y?
-- Mel Evans    spam_OUTmevans1027TakeThisOuTspamaol.com

1998\08\31@210849 by Clyde Smith-Stubbs

flavicon
face
On Mon, Aug 31, 1998 at 08:31:37PM -0400, Mel Evans wrote:
>         "The PIC 16C54 has the same performance, speed- and memory-wise, as
> the IBM-x or the PDP-y."
>     What are the values of x and y?

The PDP-8 is close, though it was slower, and had more memory (1K-4K). And all
its memory was readable and writable.

--
Clyde Smith-Stubbs               |            HI-TECH Software
Email: .....clydeKILLspamspam@spam@htsoft.com          |          Phone            Fax
WWW:   http://www.htsoft.com/    | USA: (408) 490 2885  (408) 490 2885
PGP:   finger clydespamKILLspamhtsoft.com   | AUS: +61 7 3354 2411 +61 7 3354 2422
---------------------------------------------------------------------------
HI-TECH C: compiling the real world.

1998\08\31@211023 by goflo

flavicon
face
Suggest you direct your inquiry to
.....otlistKILLspamspam.....bellsouth.net
where you may discuss the matter with the design teams
of the devices you mention...

Regards, Jack

Mel Evans wrote:
>
>     I recently heard someone say: "The PIC 16C54 has the same performance,
> speed- and memory-wise, as the IBM-709".  The UNIVAC, maybe, but the IBM-709?
>     OK, you oldtimers, here's a more general question:
>         "The PIC 16C54 has the same performance, speed- and memory-wise, as
> the IBM-x or the PDP-y."
>     What are the values of x and y?

1998\08\31@215000 by William Chops Westfield

face picon face
It's really hard to compare von Neuman architectures with Harvard
architectures, or even ram-based with rom-based architectures.

BillW

1998\08\31@220915 by goflo

flavicon
face
No ROM?

Regards, Jack

Clyde Smith-Stubbs wrote:

> On Mon, Aug 31, 1998 at 08:31:37PM -0400, Mel Evans wrote:
> >         "The PIC 16C54 has the same performance, speed- and memory-wise, as
> > the IBM-x or the PDP-y."
> >     What are the values of x and y?
>
> The PDP-8 is close, though it was slower, and had more memory (1K-4K). And all
> its memory was readable and writable.


'Oldie question: IBM-709?'
1998\09\01@013433 by Russell McMahon
picon face
>From dim and distant recall a 16C54 would give a PDP-8 quite a good
fight but the PDP had larger memory size capabilities (if you had
LOTS of money) and had awesome peripherals (paper and mag tape (to
load boot strap program from :-)) , vector graphics display, ASR33
console etc) . PDP8 could do unlimited subroutine nesting by storing
the return addresses in RAM at the start of each subroutine. (All
memory was non-volatile core so program memory was writeable). PDP 8
genuinely only had 8 instructions but had some very useful addressing
modes which allowed more powerful addressing than PIC.  The add
instruction was named TAD so you would ALWAYS remember that it was a
2's complement add - to subtract you, of course, COMPLEMENT and then
TAD. PDP8 went through quite a lot of evolution so it would depend on
the actual sub model.

One advantage that the PDP8 had over a 16C54 was that you could open
the door and walk inside it to fix it :-). We had one with an analog
addition of some sort in the engineering school in the early 1970's -
it played "Space War" (vector graphics on a Tektronix raster display,
2 ships, one moon, fired bombs with own dynamics, had gravity
even!) - you could do this with a 16C54 too.

IBM - dunno. In 1968 the university ran to an IBM BCD mini which had
to do a Fortran compile and run in 2 loads as there was not enough
memory to hold all the compiler in memory at once. You'd be pushed to
do a Fortran compiler on a 16C54 :-)



{Original Message removed}

1998\09\12@085940 by paulb

flavicon
face
Russell McMahon wrote (OK, so I«m slow in replying!):

> PDP8 could do unlimited subroutine nesting by storing the return
> addresses in RAM at the start of each subroutine.

 My first thought on that was "but what about recursive calls?".  Of
course, recursive calls are pretty uncommonly used, and *never* if the
stack depth is constrained as it most certainly is with PICs.

 Then I tend to consider; what about a utility subroutine which may
also be used within an interrupt?  That«s probably rare too.  OK then,
it«s a neat trick, FWIW.  The RETs must have been hard«coded indirect
jumps though.  (OK, I *do* have the book, I even have the FOCAL manual,
but where?)

 I«m still on the hunt for a (working) specimen too.
--
 Cheers,
       Paul B.

1998\09\12@140612 by William Chops Westfield

face picon face
   > PDP8 could do unlimited subroutine nesting by storing the return
   > addresses in RAM at the start of each subroutine.

     My first thought on that was "but what about recursive calls?".  Of
   course, recursive calls are pretty uncommonly used, and *never* if the
   stack depth is constrained as it most certainly is with PICs.

Yep. Stacks are a moderately recent invention, so to speak.  You could
check out the instruction set of DEC's PDP-10 processor familly, which
has half a dozen or so different instructions for calling subroutines.

       JSR foo         Save the PC at foo, run at foo+1
       JSP r,foo       save PC in register R, run at foo
       JSA r,foo       Save register r at foo, save PC in r, run at foo+1
       PUSHJ r,foo     use register r as a stack pointer, push the current
                       pc, run at foo.
       JRA r,foo       (return for JSA)
       POPJ r,         (return for PUSHJ)
       JRST            general purpose jump (Jump & ReStore Flags)


IIRC, the fortran 4 compiler used JSA pretty much up until the End of life
of the product line.

Ah, those were the days.  People weren't so positive that the "right"
architecture was etched it stone, so there was always something new and
weird in a processors instruction set.  Nowdays ya got yer 86 like
arhcitectures, and your RISC architectures (which are a rebellion against
the "always something new and weird" part that compiler writers really
didn't like.)

BillW

More... (looser matching)
- Last day of these posts
- In 1998 , 1999 only
- Today
- New search...