Ooooh, just two days to go! Can't you just smell the drama? (Try again. It smells a little like Lemon Pledge.) We speak, of course, of Apple's imminent music event thingy, and while we were oh-so-conspicuously not invited to the shindig itself, at least we have the option of cramming into a participating Apple retail store and catching it live via satellite like the rest of the untouchable non-media. Of course, dragging our sleep-deprived and cranky carcasses all the way out to Peabody and back just for the privilege of watching live what we'll probably be able to see later via QuickTime (if history is any indication) sounds like a total grind to us, and we'll probably skip the live feed. Must be TiVo's influence.

Anyway, as we all know, come Thursday, Apple will unleash Windows compatibility for the iTunes Music Store, in much the same way that Steve won't show up to the gig in a lime green polyester leisure suit and a rainbow fright wig. (We say that in the sincerest hope that he decides to prove us wrong. Don't forget your cameras, kiddies!) But will iTMS for Windows be too little, too late? Or even just enough too late? Reuters quotes market research analyst Rob Enderle as saying that Apple is "going to have a serious problem with the Windows community" because of the got-there-firstness of Windows-compatible copycat pay-per-download services like BuyMusic.com, MusicMatch Downloads, and I Can't Believe It's Not iTunes!™.

Now, granted, Apple may have missed a certain home field advantage by not establishing itself as a Windows-compatible service earlier, but surely it's a little early to say that there's a "serious problem." By just about all accounts, BuyMusic.com is the online music purchasing equivalent of spraying Tabasco in both eyes while beating oneself about the head and crotch with a brick wrapped in burlap and set on fire. MusicMatch has the Windows market penetration and Napster has the brand name recognition, but so far neither service offers music that'll play on the iPod, which is still "the" player to own. Maybe it's our own special brand of cynicism kicking in, but we tend to think that the Windows users who will be willing to pay for downloaded music (as opposed to going the "use Kazaa, get sued by the RIAA" route) are the ones with the cash and the class to own an iPod as opposed to one of those "other" players. So yeah, we think the iPod will likely give the iTMS for Windows a solid market base upon which to build. Frankly, we're just not all that worried. Yet.

But if Enderle's doomsayer attitude rings a bleak and gloomy bell, you've got a good memory: about a year ago, Enderle insisted that Apple would have to switch to Intel chips by the end of 2003. With barely ten weeks left in the year and the G5 kicking mucho kiester, we're going to go out on a limb and say that 1) we were right; 2) he was wrong; and 3) neener neener neener. Furthermore, we'd like to declare that anyone who predicted an all-Intel-based Apple by the end of 2003 at the end of 2002 is not equipped to make reasonable and credible predictions about any of Apple's future endeavors, because he's obviously smoking some seriously low-grade crack.

It's also interesting to note that when he made his Intel-Apple forecast, Enderle was with Giga Information Group Inc.-- but now he's apparently with "the Enderle Group," whose slogan might be "making allegedly insightful comments for over three weeks now." Suppose Giga cut him loose? Whatever. The fact that Enderle was voted "most influential analyst in the technology market in 2003" by AdWeek only underscores just how sad the technology market really is. 95% Windows market share, anyone?

Meanwhile, bring on the "unconfirmed reports," baby! We're still all a-flutter with the whispers that Thursday's media brouhaha will feature far more music-related magic than a mere launch of the iTunes Music Store for Windows. If you recall, there was talk of iTunes 5 for both platforms, with a new interface, Windows Media support, and an intriguing alleged feature called "listener loans." There may also be a bevy of iPod peripherals including video- and photo-related interfaces, a Bluetooth headset, and some sort of recording device. Somebody call Vegas and get the odds!

Well, if it means anything to you, AppleInsider adds still more "unconfirmed reports" corroborating the rumors of "major updates" to iTunes and the ability to "allow users to borrow or preview full tracks or albums for short periods of time without being billed." Needless to say, if that one turns out to be true, that's one feature that's going to be abused twelve ways from March before its feet even touch the ground. (WireTap, anyone?)

Far more interesting to us, though, is the claim that the alleged "recording device" might not be so much targeted at musicfolk, but at students: "a new iPod feature will allow students to record their afternoon lectures-- for later listening-- to an audio track via a new hardware addition to the player." Okay, that's a little vague on whether Apple is targeting students with an "iLecture" add-on or if it's just a generic voice dictation module targeted at, well, people with mouths. But if this does indeed turn out to be true and Apple is giving the iPod a little school cred, it's an effin' brilliant move-- suddenly youthful voices across the land will be lifted in a simultaneous whiny chorus of "But Moooommmmm, I need an iPod for schooooool!"

Yes, when parents look askance at the kids tossing a couple of iPods in the cart during back-to-school shopping, said kids will be able to justify the purchase by whipping out printouts of Apple-commissioned reports which show that students with recording iPods have 23% better GPAs and score 31% higher on standardized tests. Suddenly the demographic most likely to want an iPod but least likely to have the cash to spend on one will be able to rope their parents into footing the bill. Apple's iPod sales will shoot straight through the roof, and from there, it's just a hop, skip, and a jump to total global domination and the subjugation of the entire human race to its evil will. Woo-hoo!

Okay, fine, we're game; if people are going to insist on pretending that AtAT is some sort of "news source" (oooh, we feel so dirty) and then get sooooooo picky about a piddly little thing like overreporting the cost of a supercomputer by a factor of twelve, we suppose we can do a followup scene. But don't say we never gave you anything, and be advised that this is your Christmas present this year, so we don't want to hear any whining about no Chicken Dance Elmo from us under the tree come December. Maybe we'll get you a pack of Juicy Fruit as a stocking stuffer or something, but don't count on it.

Here's the skinny: that $38 million that the University of Texas is spending on its new "Lonestar" Dell-based supercomputer cluster actually covers a little more than just the Dells... which, frankly, we suspected just a teensy bit, since we were pretty sure that Dell servers-- even the dual-processor ones-- don't cost over $600,000 apiece.

Now, to be fair, if you're going to nail anyone for the discrepancy, go after the Austin Business Journal, which we cited as the source for the plot of yesterday's scene. Presumably they claim to have actual journalistic credentials and skills, but if you scope out the article on Lonestar, you'll probably agree that the reported price tag is slightly misleading. In one sentence it describes "the new 'Lonestar' computing cluster" as having more than 3 teraflops of power, and in the next it states that "the cost of the five-year project is about $38 million," implying that "Lonestar" and the "five-year project" are one and the same.

Meanwhile, there's absolutely zero mention of the fact that the "five-year project" consists of just a wee bit more than the Lonestar cluster itself-- as in, the establishment of an entire academic institute; four endowed faculty chairs to run it; funds for research and visiting scholars endowments; the construction of the entire fourth floor of a freakin' building; and a whole second supercomputer, plus the various storage systems and networking components needed to cluster it properly. Oh, and "increases in operations funding," presumably to cover an electric bill with so many zeroes that it'd make your skin fall off.

The above list comes courtesy of a UT spokesperson via The Inquirer, which also reports that the Lonestar cluster itself only cost $3 million. Okay, well, now that we have a price tag from the University itself, how about we rework that comparison with Virginia Tech's Big Mac? Lonestar cost $3 million (possibly not counting its networking and storage, but we'll let that slide) and has a theoretical peak performance of 3.67 teraflops. Big Mac cost $5.2 million and has a theoretical peak performance of 17.6 teraflops. So the Dell-based Xeon cluster is not less than a quarter as fast for seven times the price; it is, in fact, 21% as fast for 58% of the price. In other words, theoretically the G5-based solution is only, dollar for dollar, about 2.8 times as fast as the Dell/Xeon one. Our bad. We're sure that Dell and UT will be trumpeting that fact in a press release any minute now.

Yup. Any minute.