Re: triple variables?
Re: triple variables?
- Subject: Re: triple variables?
- From: Shawn Erickson <email@hidden>
- Date: Mon, 18 Apr 2005 08:59:12 -0700
On Apr 18, 2005, at 8:34 AM, Brett O'Donnell wrote:
With Tiger supporting 64-bit, is there a chance of a new variable type
that would extend the 10^306 limit of the double floating point
variable? It may sound extreme, but I'm working on a queueing
simulation app, and it requires relatively large factorials (since
170! = 10^306). It is "possible" in some cases to simplify the
algorithms to push that ceiling back a bit, but if I could just have a
G5 be the minimum requirement for numbers over 10^306 it'd be fine
with me.
I realize that the two aren't necessarily related, but with 4GB of RAM
being listed as a limit, I was wondering when if there was a way to
raise our variables to the next level too.
Review the "LP64 Data Model" section on the following page...
<http://developer.apple.com/macosx/tiger/64bit.html>
Not knowing exactly what you want in the long run consider using
vBigNum which is part of the Accelerate framework. The source for
aspect of this exists as well so you could learn from that as needed to
achieve what you want.
<http://developer.apple.com/documentation/Darwin/Reference/ManPages/
man7/vecLib.7.html>
In the end you are limited by what the hardware can do natively but you
can build on that to achieve the magnitude and precision you need.
-Shawn
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Cocoa-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden