different approach to MWSF2002
Everybody seems focused on the G5 for January 2002.
But no speculations of the G4 making another BIG clock rate step.
It might not mean too much but Motorola just silently released a new G4 version.
Also no specs on more Dual-Power-Macs or probably a Quad G4 hi-end-Mac.
I would guess more multi-CPU is not too likely but another significant step in Mhz could be possible.
Maybe even the smallest G4 Power Mac will be over 1 Ghz early next year.
If not I`d of course also be happy with the introduction of the G5 ;)
[542 byte] By [
fox congo] at [2007-11-15 6:27:34]
# 1 Re: different approach to MWSF2002
I think everyone's underestimating the possibility of a quad. It seems more technically feasable than a G5, and would be a perfect "ooh, look at that" product. And, the "arrival" of OS X actually lets those processors be useful!
Alex
# 2 Re: different approach to MWSF2002
The problem with the Quad cpu equipped machine is the available bandwidth to EACH processor. Making a motherboard that's with enough bandwidth costs A LOT to test and build.
Leonis at 2007-11-17 10:24:17 >
# 3 Re: different approach to MWSF2002
Just what we need, $6,000 PowerMacs for the masses. :rolleyes:
# 4 Re: different approach to MWSF2002
I have been for Apple to utilize multiple processors to compete (MHz Myth?) with the clock speed enhancements of the Intel and AMD world from my original registration on this board. The nearer the release of the G5, the more remote the possibility that Apple would offer MP in any major way.
Having said that I'd love to see duals & quads offered, If Motorola has a large supply of G4's on hand and they try to unload them before the release of the G5. One can hope...
# 5 Re: different approach to MWSF2002
quad macs processor macs exist. DayStar made a quad 604e 180Mhz machine back in the day. from what I understand it performed very well.
I was very close to buying one but opted for a dual PowerMac 9500 instead... what can I say.. I like the apple logo :)
imagine BeOS on a quad 604e... drool
# 6 Re: different approach to MWSF2002
Imagine OS X (kernel compiled to take advantage of more than 2 processors) on a Quad G4 7640 at whatever MHz. :eek:
[ 11-25-2001: Message edited by: macway ]</p>
macway at 2007-11-17 10:28:26 >
# 7 Re: different approach to MWSF2002
Apple should just announce outragous product and see how Dell copies 'em
word :cool:
# 8 Re: different approach to MWSF2002
I thought the current G4 wasn't fully MERSI compliant and thus could only go 2-way MP and not 4-way. Can someone verify this? (maybe this is why we saw a silent update by Motorola... fully MERSI compliant now?)
# 9 Re: different approach to MWSF2002
[quote]Originally posted by crawlingparanoia:
<strong>I thought the current G4 wasn't fully MERSI compliant and thus could only go 2-way MP and not 4-way. Can someone verify this? (maybe this is why we saw a silent update by Motorola... fully MERSI compliant now?)</strong><hr></blockquote>
I think the fully MERSI compliant only matters in configs of 8 or more or something like that.
I think the bigger problems are
1.) bus
2.) the need
3.) the price
4.) the market
5.) r/d costs
# 10 Re: different approach to MWSF2002
MERSI compliance (as implemented in the original 7400 G4) allows a processor to support up to 8 way SMP without additional support logic (ie, circuitry that is dedicated to making sure that all the CPUs are working with the right data).
MESI (7450, 604) compliance leaves out a small but crucial bit that the CPUs use to ensure cache coherence between them (in short, that they're all working with the same, current data: otherwise, one processor might have an older value for a given variable than another one). That omission limits the number of symmetric processors to 2 before support logic is needed..
MEI (603, 750) compliance means that support logic is required for any kind of SMP.
Amorph at 2007-11-17 10:32:32 >
# 11 Re: different approach to MWSF2002
If i'm not mistaken, the 7460 G4 will re-instate MERSI compliance...
:D
