Re: [Phys-L] I need big numbers crunching

["John Clement" <clement@hal-pc.org>]

I used an IBM computer in the '70s that was a decimal machine.  It did all
arithmetic as BCD decimal in hardware.  You could specify the length of the
computer word for integer or floating point arithmetic, and it was built
into the hardware.  You specified the number of digits to be used in the
JCL, and it could be almost any arbitrary length.  But it was slooooow and
had very small storage.  You could look at the lights and almost see it
doing its operations.

I think it will be hard to get a machine with anything other than the
standard size arithmetic words, especially in floating point.  You may be
able to find one with arbitrary length fixed point.  Perhaps you should
contact someone who does things like calculating PI to humongous precision.
With standardization of computer precison and architecture you pay the price
of being able to easily do some diverse things.  For example I used a
machine where you could redefine standard operators like *,+,-... to do
completely arbitrary functions, but that is probably not common now.

John M. Clement
Houston, TX

> -----Original Message-----
> From: Phys-l [mailto:phys-l-bounces@phys-l.org] On Behalf Of 
> Roberto Carabajal
> Sent: Saturday, March 22, 2014 11:17 AM
> To: phys-l@phys-l.org
> Subject: [Phys-L] I need big numbers crunching
>
> Hello:
>
>  
>
> Please, can you recommend a programming languaje 
> implementation such can deal with arbitrary real numbers 
> length (no Maple nor Mathematica) in a simple way (native) ?