|
|
Magic Mailing List |
|
From: erik peterson (erik AT yanntek DOT com) Date: Fri Sep 28 2001 - 11:14:42 EDT
Richard Lethin wrote: > > "R. Timothy Edwards" wrote: > > it's really not proper to assume that everybody "knows" that you > > have to write CIF, read it back in verbatim, and check for these kind > > of violations. > > I guess you need to check somehow for connectivity violations > (resimulating or extracted > netlist comparison) and new DRC errors. More broadly, are there any > pointers to documentation on foolproof MAGIC tapeout procedures online? > I've been putting one together but it's all hard-learned errors. It > would be nice to fortify it with someone else's checklist. > > Thanks, > > RIch i havent made any documentation suitable for outside my company on this check, but we are very sensitive to this issue. our generic tapeout flow is simple: (there are exceptions for every project) lvs: magic -> 'layout.strm' -> space -> 'layout.net' spice -> 'spice.net' 'layout.net'+'spice.net' => gemini --> tools for lvs: 1. netlist extraction: 'space' http://cas.et.tudelft.nl/space/space_soft.html 2. netlist compare: 'gemini' http://www.cs.washington.edu/research/projects/lis/www/gemini/gemini.html you can catch a lot of 'cifout' section bugs through lvs. space comes with a viewing tool (dali) which comes in handy when you are building the cifout section. we also use multiple magic techfiles for one process when there are errors that are impossible to check with only one pass (like antenna rules, checkerboard, etc..). late for work so i have to cut this short. space installs in minutes, you can build the most basic techfile in half a day (cold start) and use the viewer to doublecheck gross cifout errors (yes it has limited DRC). -erik
|
|
|
|