Xref: utzoo comp.arch:19085 comp.lsi.cad:709 Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!zaphod.mps.ohio-state.edu!sol.ctr.columbia.edu!cica!news.cs.indiana.edu!ux1.cso.uiuc.edu!ux1.cso.uiuc.edu!aglew From: aglew@crhc.uiuc.edu (Andy Glew) Newsgroups: comp.arch,comp.lsi.cad Subject: >32 bit integers Message-ID: Date: 8 Nov 90 17:42:24 GMT Sender: news@ux1.cso.uiuc.edu (News) Followup-To: comp.arch Organization: Center for Reliable and High-Performance Computing University of Illinois at Urbana Champaign Lines: 30 Another data point for the low-heat comp.arch discussion of "Are addresses/integers >32 bits necessary?": In article "Multichip modules test design tool limits" November 1, 1990, Computer Design, pg. 68, quoted without permission: "``So that the fine pitches of MCMs could be dealt with, we adapted and reworked our Allegro circuit board tool. We changed the database, for instance, from a 32 bit integer capability to a double-precision floating-point, 64-bit design to handle the finer pitches''" quoting Shiv Tasker, director of product marketing in the printed circuit board division of Valid Logic Systems. From my limited knowledge of CAD tools, I believe that the floating point values are being used to preserve coordinates on a grid. Ie. they are used for exact, integer-oriented, calculations. I conjecture that it would have been preferable to use 64 bit integer values if these "long long" integers had a performance advantage over floating point. (Of course, now that Valid has done the conversion, it wouldn't be worth converting back unless the performance gain was big). -- Andy Glew, a-glew@uiuc.edu [get ph nameserver from uxc.cso.uiuc.edu:net/qi]