[SI-LIST] : Trusting tools.

Laurence Michaels (lmichael@techmail.gdc.com)
Thu, 15 Jul 1999 16:48:56 -0400

In the 'trusting tools' vein...

At least one tool I have used is worthlessly inaccurate.

The software vendor finally admitted that they knew the tool didn't work
for the sort of application I was trying. In my estimation, my
application was something that *should* have worked, and if it didn't,
then any other result from that tool was suspect. I no longer use that
tool, and am seriously considering switching away from some of their
other simulation tools.

I have found it very difficult to prove whether or not a particular
simulation is accurate. It is also difficult to assess how accurate a
sim might be. Simulating digital systems is not as hard as simulating
the SI characteristics of those systems.

I can think of several reasons why that might be, including 'digital
systems are used to simulate themselves' for digital simulation, and
'discrete systems have difficulty simulating continuous events.'

I suppose the only way to be sure that a tool works most of the time is
to run multiple test cases on boards similar to the ones being designed.

Regards,

-- 
Laurence Michaels
Any technology distinguishable from magic is insufficiently advanced.

**** To unsubscribe from si-list: send e-mail to majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE si-list, for more help, put HELP. si-list archives are accessible at http://www.qsl.net/wb6tpu/si-list ****