Book E standard, and non-compliant PPC970s
Ti_Poussin said:
I'm not sure what you're talking about, but maybe you can light me up on this one (it's not a joke or sarcass).
Fair enough....
http://www.eetimes.com/story/OEG19990507S0003
IBM, Motorola write Book E on the PowerPC
David Lammers David Lammers
EE Times (05/07/1999 11:02 AM EDT)
SAN JOSE, Calif. Motorola Inc. and IBM Corp. have announced Book E, a jointly written architectural definition and instruction set for embedded 64-bit PowerPC implementations.
Book E may give Motorola and IBM a way to present a common front in both the embedded and desktop sectors. It will build a foundation under system-on-a-chip designs that might incorporate intellectual property (IP) cores sourced from commercial IP vendors, or from the core libraries now under construction at both IBM and Motorola.
Over the next year both companies will come out with next-generation PowerPC controllers that adhere to the Book E definition, and will work to ensure that tool vendors create a common software-development environment. The companies also will offer PowerPC licenses to customers and foundries that want a wider number of sources for high-volume designs."
See also
http://news.com.com/2100-1040-225442.html?legacy=cnet
So, Moto and IBM create an architectural definition of what a PowerPC should be, and encourage vendors to "follow the book".
http://encyclopedia.lockergnome.com/s/b/PowerPC (a page or two down, under the heading "Design Features"):
"The PowerPC is designed along RISC principles, and allows for a superscalar implementation. Versions of the design exist in both 32-bit and 64-bit implementations. Starting with the basic POWER specification, the PowerPC added:
Support for operation as in both Big-Endian and Little-Endian modes; the PowerPC can switch from one mode to the other at run-time (see below). This feature is not supported in the PowerPC G5. (This was the reason why Virtual PC took so long to be made functional on G5-based Macintoshes.)"
The next couple of paragraphs explain it in more detail.
See also
http://www.shahine.com/omar/CommentView,guid,e706c1bf-87fe-4cd0-aa78-a78ac023cd4a.aspx
__________________________________
So, there was all the hype about a "common, compatible architecture", and Connectix did the work to use one of these defined features to significantly improve VPC's performance.
Then, Apple stupidly (because they didn't realize it) or arrogantly (because they didn't care) chose a non-Book E compliant chip from IBM - the PPC970.
This not only broke Virtual PC for quite some time, but reduced the performance because of the non-standards compliant CPU that Apple chose.
But of course, you often find comments around here that it was something that Microsoft deliberately did to hurt the Mac.
You can also search the web for
"virtual pc" g5 "little endian" for lots more on the topic.