home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.sys.mac.system:15337 comp.sys.mac.hardware:24502 comp.sys.mac.advocacy:3400
- Path: sparky!uunet!stanford.edu!apple!tim
- From: tim@Apple.COM (Tim Olson)
- Newsgroups: comp.sys.mac.system,comp.sys.mac.hardware,comp.sys.mac.advocacy
- Subject: Re: PowePC Chip emulation, Why 68020??
- Message-ID: <75694@apple.apple.COM>
- Date: 17 Dec 92 16:41:33 GMT
- References: <1992Dec16.073119.1376@galaxy.gov.bc.ca> <zkessin.724522396@chaos.cs.brandeis.edu>
- Followup-To: comp.sys.mac.system
- Organization: Apple Computer Inc, Cupertino, CA
- Lines: 27
-
- In article <zkessin.724522396@chaos.cs.brandeis.edu> zkessin@chaos.cs.brandeis.edu (Zach, the Lego Maniac) writes:
- cconstantine@galaxy.gov.bc.ca writes:
-
- > Does anyone out there know why Apple chose to emulate a M68020 for
- > its' PowerPC chip for software compatability instead of at lease a
- > 68030 which can do virtula Memory and was the most stable chip (as far
- > as software compatability went) of the bunch????
-
- |Prob. because a 68020 can do virtual with a 68851 chip, which will
- |prob. also be emulated. as will the coprossor. Plus the mac's memory
- |will have to go threw the rs6000 memory manger.
-
- If you think about it, emulation of a 68K MMU (the 851 or the MMU on
- the 68030) would be too much of a performance hit -- each access would
- have to go through software emulation of the protection and
- translation hardware. There is no reason to do this, as:
-
- 1) The operating system can work without this functionality
- 2) The PowerPC processors have their own MMUs
-
- Other than the addition of the MMU and a data cache, the 68030 is
- essentially equivalent to a 68020...
-
- --
- -- Tim Olson
- Apple Computer Inc. / Somerset
- (tim@apple.com)
-