Decimal V Binary??

[Bjørn Vermo]

On Thu, 23 Sep 2004 07:40:20 -0400, Jim Donoghue <jim at smithy.com> wrote:

> On Wed, 2004-09-22 at 22:08, William Donzelli wrote:
>> > There are SOME people (I am NOT one) who actually feel that
>> > DECIMAL is the right way to go in computers!
>>
>> If the computers are doing my bank statements, then yes, I am one of  
>> them.
>
> I've always wondered what systems/processors (other than IBM mainframes)
> that had hardware decimal arithmetic support. I don't mean 'decimal
> adjust' instructions, I'm talking about full
> add/subtract/multiply/divide/etc. The only other one I know of is the
> Wang VS, which came standard with support for 'packed decimal' data, and
> even floating-point decimal. The Wang VS instruction set was very
> similar to the IBM 360/370, even the packed decimal data format was the
> same.
>

If I remember right, the NCR315 had full BCD arithmetic. At least the  
arithmetic instructions it had were decimal, I do not remember if it  
actually had a divide instruction in hardware.

It was even so decimal-oriented that memory was addressed in decimal.

A rather nice feature was that all internal registers were ferrite core.  
If it stopped because of a power falure, it could be restarted in the  
exact instruction step where it had stopped.

When I was working for NCR Norway in the early 70's, the bulk of input to  
the 315 came from OCR - reading cash register rolls printed in National  
Optical Font. Not bad for a cumputer which was shipping earlier than the  
IBM 360.

We also had dial-up data transfer over modem from England, but that seemed  
a bit trivial when Norway was connected to ARPANET the year after. That,  
however, is a totally different story.

-- 
Using M2, Opera's revolutionary e-mail client: http://www.opera.com/m2/