You know, if you're starting to harbor some lurking doubts about this whole IBM-making-G5s thing, it's okay to admit that. Sure, back when the G5 was first announced, Big Blue came across as the Mac platform's knight in silicon armor; it was only natural, seeing as we'd all just slogged through years of Motorolan chip stagnation (for about a year and a half, there, the company's rallying cry appeared to be "half a gigahertz is plenty for everybody") and processor supply droughts, some of which resulted in Mac shortages and ensuing earnings warnings that smacked Apple's stock price down so fast most of us got nosebleeds from the sudden pressure change. And then along comes IBM, twirling a 2.0 GHz PowerPC that spanks the G4 in every conceivable way (well, except maybe for power consumption), and there's Steve, promising a 3.0 GHz Mac within twelve months. Of course we all hoisted IBM on our virtual shoulders and dunked its proverbial head in the metaphorical Gatorade. Figuratively speaking, of course.

But let's face it: while the G5 is clearly one spankin' chip, IBM's delivery of it hasn't exactly been stellar. The original G5s shipped late (okay, we can partially blame Virginia Tech for that), 2.0 GHz processors were scarce enough that Apple eventually introduced a dual 1.8 GHz model to stem some of the demand, and the new 90-nanometer chips were late enough to postpone the Xserve G5's ship date by a month and possibly torpedo those 2.5ish GHz Power Macs the rumor mill has been expecting since the beginning of the year. None of this has exactly bolstered confidence in the prospect of hitting 3.0 GHz within the next three to five months. Indeed, bits of the saga sound almost Motorolesque, lending credence to the theory that a curse dogs whatever company dares to manufacture chips for Apple's highest-end systems.

To tell you the truth, the Curse Theory is starting to make more and more sense to us these days. IBM is definitely not shipping as many G5s as it's supposed to-- Apple confirmed that in its last quarterly earnings conference call-- and yet Motorola (or, to be perfectly accurate, Freescale, who wanted to call itself "The Chip Company Formerly Known as Motorola's Semiconductor Division" before it discovered that Prince already has the trademark registered) seems to be recovering somewhat. Faithful viewer Jef Van der Voort dished us a recent MacRumors blurb which noted that Freescale's PowerPC roadmap now includes a mention of the G4's non-IBM successor, the e700, which will run both 32- and 64-bit software and will scale "to 3 GHz and beyond in next-generation process technologies."

So, keeping The Curse in mind, any bets on which company will make it to 3.0 GHz first? Answer: presumably whichever company isn't making them for Apple's Power Macs. It's like lines at the supermarket; whichever checkout lane you move to winds up becoming the slowest. Seriously, folks, we need to find someone who can lift this curse before we're all looking longingly at 3.2 GHz Freescale e700s next year and IBM is still saying "3.0 GHz coming up any minute now." Anyone here adept at the black arts? Anyone? Bueller?