home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.ibm.pc.misc
- Path: sparky!uunet!ukma!darwin.sura.net!convex!news.utdallas.edu!goyal
- From: goyal@utdallas.edu (Mohit K. Goyal)
- Subject: Re: Can software be 486-specific?
- Message-ID: <BxMEqn.C7A@utdallas.edu>
- Sender: usenet@utdallas.edu
- Nntp-Posting-Host: csclass.utdallas.edu
- Organization: Univ. of Texas at Dallas
- References: <BxL9qH.M45@utdallas.edu> <RFowler.393@kelvin.jpl.nasa.gov>
- Date: Thu, 12 Nov 1992 20:39:58 GMT
- Lines: 19
-
- >Remeber that the 2 or 3 opcodes that are different between a 386+387 and a
- >486 are for MULTITASKING OPERATING SYSTEMS, and MULTIPROCESSOR support.
- >Therefore, it kind of makes sence for a new 32bit operating system (
- >NeXTSTEP) to require an i486 processor. Windows NT, by the way, requres a
- >386-25 or grater, not a 486.
- > _________________________________________________________________
-
- How much difference(ie-improvement) do these 2 or 3 opcodes make in a
- multitasking operating system? How much better would an OS like OS/2
- be if it required a 486SX instead of 386SX?
-
- Finally, how can NT require a 386-25 or greater? Can't an OS just
- require either a 386SX & up, 386+387+up, 486SX+up, 486DX+up?
-
- Does MS just recommend a 386DX-25 as the absolute minimum?
-
- Thanks.
-
-
-