To: Al. Gibberd who wrote (9269 ) 3/11/1998 4:53:00 AM From: RyanF Read Replies (3) | Respond to
In regards to the Merced situation, I am under the impression that the systems are initially going to be targeted at the workstation market. I remember an article stating that the entry point for a Merced machine would be 8-10k. Which would make sense since the chip is going to cost $1500-2000, on par with an Alpha chip. Now I really don't think that the G4 is going to be selling at this price since it will need a much smaller wafer size, which is (if I'm not mistaken) the major overhead in chip fabrication. Economy of scale might be a factor, but I doubt it. Another extremely important factor no one has raised is that Merced is an entirely new chip design, VLIW, which will entail a major rework of software. Granted, Merced will contain a compatibility core which will allow the execution of x86 code, but it seems to me that is BLOAT-CHIP == INTEL / BLOAT-WARE == MICROSOFT. It is interesting to contrast the 68000->PPC and x86->Merced transitions. In the first case, the benevolent dictator (Apple) wrote a emulation layer on the MacOS. Creating hardware emulation on PowerPC would have been blasphemous to the whole RISC religion. Plus of course, it made no sense putting the chip design in crutches just to satisfy a customer (Apple) that was facing increasing marginalization. On the other side we have Microsoft which is Intel's bread and butter. So it must throw in some x86 hardware into the Merced design to satisfy Bill. Intel however resents Microsoft's monopoly over its bottom line and starts shopping for other customers, namely the assorted UNIX vendors. Intel even sends over a dozen engineers to Be in order to make sure the porting process goes without hitch. Which by the way, allows the BeOS to run faster on Intel than PowerPC, snail-ad be damned. In my opinion, Merced will be a smashing success simply because Intel is one smart cookie. Motorola gave up on the 68000 series. I can't imagine the mental effort and creativity it must take to keep the x86 apace of the PowerPC even though they are separated by something like 8-9 chip generations (Moore's law). It's always easier to throw away the old ideas. PowerPC has always seemed unfocused to me. All three members of the alliance have varying amounts of interest in the promotion of the PowerPC. Intel of course has one goal: making, promoting, and selling its next best chip. And as the tenuous technological lead of copper wire design melts away, Merced's superior ideas and the raw talent behind it will enable it to rise to the fore, IMHO. Where does Apple stand? Hopefully with one foot in Lake Merced and the other in PowerPC Pond. Luckily the pond has been recently cleared of the leeches that were sucking poor Apple dry. One of the keys to Apple's future is in their flexibility, which is what Rhapsody affords them. I'm feeling a little verklempt...Discuss amongst yourselves.
