Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!wuarchive!gem.mps.ohio-state.edu!samsung!xanth!mcnc!duke!helios!vishin
From: vishin@helios.cs.duke.edu (Sanjay Vishin)
Newsgroups: comp.lsi.cad
Subject: Re: Where do I get Magic, etc. for Sun4
Summary: crystal performance
Message-ID: <15919@duke.cs.duke.edu>
Date: 30 Oct 89 02:39:45 GMT
References: <91600002@uicbert.eecs.uic.edu> <6124@pitt.UUCP>
Sender: news@duke.cs.duke.edu
Lines: 24

In article <6124@pitt.UUCP>, shorty@jupiter.cs.pittsburgh.edu (Saverio Fazzari) writes:
> 	I want to determine delay times for them, right now I am using
>         crystal but a extraction of resistance from magic is not very good
>         what is the best way to do it , or is there a better tool out there?
> 				shorty@ee.pitt.edu

 I have seen people complain about "Crystal" before. Agreed that it does 
 estimate delays in a pessimistic manner, but it still has the capacity to
 "differentiate between two delay paths", which is more important for a delay
 /timing analyzer, than accurate prediction of delays. Once the time
 consuming process of weeding out the main (say 5 major) delays paths has been 
 done, the rest (accurate delay estimation) can be done by a circuit simulator 
 like spice (which Crystal produces automatically).

 In the same context, resistance estimation by Magic may not distort the
 delay results as much as crystal itself would using its rough RC estimation
 model. Does anybody have any results on this ? What about TV (does anybody 
 use that) ?
 

 Sanjay Vishin-

 vishin@cs.duke.edu
 vishin@mcnc.org