Re: 100-points limit discussion

From: D. C. Sessions <dc.sessions@vlsi.com>
Date: Thu Jun 10 1999 - 16:03:44 PDT

"Muranyi, Arpad" wrote:

> 6) On the other hand, if we do not care about SPICE PWL sources, we can
> raise the
> limit to practically any number. The things we need to consider then are
> what some
> already brought up in this discussion, file size, distribution, but I would
> add
> speed of simulation also. My understanding is that the larger the tables
> get, the
> more if-then-else evaluations will need to be done which can slow down the
> simulator.
>
> 7) I tend to agree that unlimited points may not be a good idea, because
> (forgive me)
> there is no cure for laziness and stupidity. We may end up with a lot of
> large files
> with no reason, or added accuracy.

I would *hope* that nobody is using raw PWL for modeling. Not only is
it slow and bulky, but worst of all it's unnecessarily inaccurate.
Cubic spline algorithms are fast and only need to be run once (when the
model is read in) and can even be used with simple scripts supporting
SPICE simulators. After that you use a piecewise-cubic model with a
minimum of inflection points and NO first- or second-order discontinuities.

The large number of source points is needed mainly to make sure that
there is plenty of coverage in 'interesting' regions, and secondarily
as a mechanism for judging when to remove inflection points.

-- 
D. C. Sessions
dc.sessions@vlsi.com
Received on Thu Jun 10 16:10:45 1999

This archive was generated by hypermail 2.1.8 : Fri Jun 03 2011 - 09:52:30 PDT