Everyone knows the drill by now: Apple could announce an inexpensive plan to cure world hunger, and CNET would step up and condemn it for not using the right brand of ketchup. There is seemingly no extent to which CNET won't travel in order to take a positive development out of Apple, and spin it negatively. This is, after all, the outlet that tried to make the case a little while back that iPod demand was "iffy," and every few months, runs the exact same "Five reasons not to buy an iPod" article on its front page. But in one of the most shameful acts of journalistic larceny I can remember, CNET ran an article back in January (not surprisingly, with no author listed) that used a variety of ridiculous tactics in order to attack Virginia Tech's supercomputer, and to attempt to leave the reader thinking that what Apple and Virginia Tech managed to pull off together wasn't real.
Here's a brief history of the project, if you missed it the first time around: Virginia Tech decided to construct a world-class supercomputer on its campus, and after looking at other alternatives, concluded that it could simply purchase eleven hundred off-the-shelf PowerMac G5 computers, string them together, and have the kind of computing power it wanted at only a small fraction of the cost of any other cluster ever created. It's clear that there was no pro-Macintosh bias involved in the choice to go with Apple, when you consider that the guy who decided to go with Apple, wasn't even himself a Mac user. And it worked; not only did the whole project cost a mere $5.2 million (this is apparently chump change in this arena), it immediately became the third-most powerful supercomputer in the world. In other words, such a rousing success that scientists from all over the world began flying into Blacksburg just so that they could get a good look at it and steal the idea. There was really nothing negative that could be said about the project...and yet CNET went about doing just that.
Since I hope that you've got better things to do than read the entire CNET article (I've linked to it merely to show that it really did exist), I'll sum it up for you. Rather than recognize the fact that building a Mac-based cluster is so easy that Virginia Tech literally convinced some of its undergrads to assemble it in exchange for free pizza, CNET claimed that the actual cost of the product was much higher than reported, because when calculating the project's cost, Virginia Tech failed to include the hourly wages that they didn't pay to the students. Or perhaps it was that Virginia Tech failed to include the cost of the pizza. I can't rightly remember.
But the article gets even worse. In its last paragraph, instead of recognizing that Apple and Virginia Tech had in fact set the supercomputing world on its ear, and that it was all but guaranteed that Mac-based copycat clusters would pop up in other places before long, CNET instead attempted to explain how this was nothing more than a fluke that would never be repeated. Their words, not mine: "customers who have less interest in experimentation are more likely to pay specialists at Linux Networx, RLX Technologies, IBM, Sun Microsystems, Dell or Hewlett-Packard to plan the plumbing, package the software and plug in the cables. And those companies aren't going to rely on Macs."
Wowzers. Sounds more like a desperate plea than a prediction. Whoever wrote the January article (as I said, no author is listed) was certainly hoping that no one else would follow in Virginia Tech's footsteps. But guess what happened last night? Word leaked out that a certain someone had purchased no less than fifteen hundred Apple XServes (a custom version of the off-the-shelf Power Macintosh G5 built specifically for stacking them on racks), and plans to use them to build a supercomputer cluster of their own. The architect of the cluster is a little company called COLSA, which you've never heard of. But the client that COLSA is building the cluster for just happens to have a bit more name recognition...perhaps you've heard of them: the United States Army.
That's right kids, the Army wants a supercomputer, and by "late fall," it'll have one. It will be so powerful that it just might rank as high as second-most powerful in the world. And it'll be entirely Macintosh-based. But other than the fact that it's the Army that's doing so, this isn't surprising at all. It was just a matter of time before someone came along and out-did Virginia Tech by using the same Mac-based blueprint but one-upped them by using even more Macs. So the only real question now, is this: how exactly is CNET (or anyone else) going to manage to negatively spin this one?
And the answer is that I don't think they can. I don't think anyone can. And for that matter, forget about CNET, the real question is whether the general public can manage to dismiss this one. Because the real issue here is not just about how many more Mac-based copycat clusters manage to spring up. What I'm interested in is just how many folks sitting at home will hear this news and finally begin to "get it" when it comes to the fact that they'd be better off with a Mac in their home than their current Windows PC. Is it rational to switch computing platforms in your home based on what the Army is doing with cluster computing, rather than basing it on what's best for your own needs? Of course not. But as evidenced by the fact that the vast majority of people are currently using Windows, consumers are clearly using some rather nonsensical methods for determining what to use at home, as it is. So for once, why not some nonsense that steers them in the Mac's direction?
You might wonder why the Army's Mac-based supercomputer might have the kind of influence over Joe Six-pack that Virginia Tech's cluster didn't. Well, for one thing, this is the second time around, and proves that the first wasn't just a fluke. It shows that CNET's concerns about hidden costs were unfounded, unless you happen to think that COLSA corporation pays its employees' salaries in pizza instead of cash. Clearly, the company weighed in the costs of assembly, and still concluded that an Apple-based solution as the cheapest. In fact, the Army's cluster, despite employing 50 percent more machines than Virginia Tech (1556 vs. 1100), is only projected to cost about 10 percent more money ($5.8 million vs. $5.2 million). And, depending on who you believe, the Army's cluster will be anywhere from a mere 50 percent (according to CNET, not surprisingly) to an astounding 150 percent (according to MacCentral) more powerful than V-Tech. So it looks like the cost of building a Mac-based cluster is going down, not up.
Clearly, having a second Mac-based supercomputer cluster come into existence in as many years bodes well for the credibility of such a project. But while that's not likely to get your average consumer racing to their local Apple Store to switch platforms, here's what might: this isn't a college, it's the Army. In my time, I've run into plenty of parents who think it's great that schools use Macs (and even agree that the Mac is the better computer and want to see their kids' school using them), but when it comes down to it, they still want a Windows PC in their home, because it's what the "real world" uses (whatever that might be). And no matter how you slice it, Virginia Tech is undoubtedly a school. In other words, it's far too easy for the average consumer to take one look at V-Tech's Mac-based supercomputer, and write it off because "schools always use Apple."
But the Army? That's no school, folks. And while military life may be even further removed from the "real world" than college life, no one sitting at home is going to think that the Army went with Macs because of any pro-Macintosh bias. In fact, with the Army being a government agency, its decision to go with a Mac-based solution is probably even more likely to get people's attention. Sure, NASA is chock full of Macs, even though no one in the "real world" seems to be aware of it. But after Steve Jobs gets done devoting a good chunk of his WWDC Keynote to bragging about the fact that the Army has just purchased 1500 Macs, rest assured that the mainstream press will pick up on it and share it with those folks at home who don't drive to their local Apple Store just to watch a live video of every Steve Jobs keynote.
And this one will be darned near impossible to spin even remotely negatively. Even CNET can't seem to come up with anything negative to say about it, instead referring to it as the "real-world cousin" of of the Virginia Tech cluster. It would appear that even the most consistently anti-Apple-biased news outlet out there is stumped as to how to dismiss Apple's achievements as irrelevant this time. So when the legitimate news outlets (who, despite not knowing nearly enough about the Macintosh, could likely care less about intentionally trying to paint Apple in a negative light) get ahold of this, it ought to come off as a rather bulletproof endorsement of the Mac platform...at least in the minds of those sitting at home on their couches, convinced that the Mac isn't used in the "real world."
True to form, CNET just couldn't resist taking an intentionally misleading pot-shot at the Virginia Tech cluster, referring to it in the sub-title of the article as having "briefly appeared high on a list of the most powerful machines." CNET is clearly hoping that you'll interpret that to mean that the cluster wasn't powerful enough to remain on the list. But in reality, Virginia Tech is currently in the process of retrofitting its cluster, swapping out its original off-the-shelf PowerMac G5's with the same G5 XServes that the Army is using, which didn't exist at the time that Virginia Tech originally constructed its cluster last year. So the only reason that V-Tech's cluster isn't currently on the list of the top supercomputers, is that it isn't currently turned on. Nice try, CNET. I don't know what you're trying to accomplish, but all you're doing is showing people like me that you can never, ever be trusted to report anything in a straightforward manner.
But here's the cool part: despite the fact that Virginia Tech's cluster wasn't on the most recent list (due to being turned off for retrofitting), and the fact that the Army's cluster wasn't on the list either (because it hasn't been assembled yet), both machines are expected to be up and running by the time the list comes out again in November. And at that time, two of the world's four fastest supercomputers will be entirely Macintosh-based.
Will CNET still attempt to refer to Apple as "far from prevalent in the supercomputing arena," as they did in January? I don't much care. The real question here is whether the United States Army's construction of a Mac-based supercomputer manages to do anything to help dispel the popular myth that Macs are never used in the "real world." For the life of me, I don't know what the "real world" is supposed to be, but even CNET appears to agree that this qualifies.
And with that, I think that my desire to see Steve Jobs' Keynote on Monday just went up by a whole-number multiple.
hey folks, click here to shop Amazon and support us...
