home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!pipex!warwick!uknet!bilpin!nuntius
- From: mark@bilpin.co.uk (Mark Allerton)
- Newsgroups: comp.sys.mac.hardware
- Subject: Re: Motorola's 68060 for the Mac
- Message-ID: <1993Jan21.141259.6424@bilpin.co.uk>
- Date: 21 Jan 93 14:12:59 GMT
- References: <1993Jan20.030348.20917@bilby.cs.uwa.edu.au>
- Sender: news@bilpin.co.uk (NetNews Administrator)
- Organization: UK Mac DTS/SRL Data
- Lines: 31
- X-Useragent: Nuntius v1.1
-
- In article <1993Jan20.030348.20917@bilby.cs.uwa.edu.au> Quinn,
- quinn@cs.uwa.edu.au writes:
- >o The PowerPC is a real 64 bit chip with 64 bit registers, 64 bit data
- > paths and (I assume) dual issue of 32 bit instruction words. A nice
- > chip.
-
- My understanding is that the PowerPC is in fact a 32-bit CPU, with
- 32-bit registers, that can be scaled up to 64 bits, and that in a
- 64-bit version (*not* the 601) then programs could switch between
- 32-bit mode (the default, most likely) and 64-bit mode. This
- switching will be available as a user (rather than supervisor) mode
- instruction.
-
- The way I had it explained to me (by a man from Motorola) was that
- most applications would probably work in 32-bit mode (32-bit ints
- being pretty much all you need most of the time) but might switch
- to 64-bit mode for stuff that really needs it (bitmap operations,
- maybe?)
-
- This is all pretty consistent with what I read in the PowerPC
- article in Microprocessor Review which was published on the 1992
- WWDC CD-ROM.
-
- I haven't a clue how they plan to deal with the issues that arise
- from going to 64-bit addressing though.. (i.e sizeof(void *) !=
- sizeof(long))
- but those problems might go away to an extent if each process has a
- seperate address space starting at zero...
-
- Mark Allerton
- SRL Data, but I speak for myself
-