Fun fact: believe it or not, folks, AtAT's wild success isn't confined to these here United States. No, seriously, it's true! The show actually has semi-regular viewers holed up in such far-flung corners of the world as Iceland, the Dominican Republic, and Delaware-- and for the benefit of those fans, we thought we'd explain that, here in the U.S., today we celebrate a holiday called Columbus Day. Columbus Day, for the uninitiated, is one of our most sacred occasions: a day on which we reflect on the many cultural and technical achievements of the city of Columbus, Ohio. We celebrate Greater Columbus's world-famous quilts, its shrubberies recreating Pointillist masterpieces, and (most importantly) its commitment to the preservation of really old TV sets by wondering why the bank is closed and our mail never came. A good time is had by all.

So if this is such a major holiday, why are we broadcasting, you ask? Well, normally we wouldn't, but faithful viewer Nathaniel Madura pointed out that Slashdot just referenced a BBC World report on that G5 supercomputer down at Virginia Tech, and we're just a little giddy about the existence of a Mac-based cluster than can chew through 17.6 trillion floating point operations per second. Yes, the thing is up and running (at least enough to run performance testing), and reportedly it pumps out 17.6 teraflops of raw perforated aluminum muscle while sucking down enough juice to power 3,000 average homes. Wow, is it getting warm in here, or is it just us? (It's just us-- the G5s are cooled by means of 1,500 gallons of chilled water pumped through every minute. Ooooo, frosty.)

Kudos to the Virginia Tech team who pulled this off, because frankly, this is the sort of technological triumph we'd normally only expect to come out of, say, Columbus, Ohio.

Now, what's interesting about that 17.6 teraflop figure is that if you scope out the last compiled list of the world's top 500 supercomputers (from last June), you'll notice that, if 17.6 teraflops is Virginia Tech's "theoretical peak performance" score, it'll probably slot in nicely at number three. (Scores are ranked by "Maximal LINPACK performance achieved," so it's just guesswork so far.) The top-ranked Intel-based cluster is currently ranked at number three, with 2,304 2.4 GHz Xeons and a theoretical peak performance of 11 teraflops. Gee, more processors, a higher clock speed per processor, and 63% of the performance. Now that's efficiency, baby!

We'll have to wait until the next top 500 list comes out in November to see if "Big Mac" (as the VA Techies apparently call it) really takes third place, or if the real-life LINPACK scores push it down lower-- but we figure a top five placement is a safe bet. One of the world's bestest supercomputers made of Macs and running Mac OS X? Why, it's a Columbus Day miracle!

Meanwhile, we know that the G5 supercomputer is delivering more pluck per processor than any other supercomputer out there, but what about bang for the buck? Well, remember how Virginia Tech announced that they went with G5s because they were so much cheaper than Dells? Now we've got something amounting to proof: faithful viewer Pedro Henriquez notes that Michael Dell had yet another one of his psychotic "Gotta Be Steve" episodes recently and had a Dell supercomputer built at a big university. Surprise, surprise. The only problem is, he didn't copycat the move quite enough to deliver anything close to the same degree of raw performance-- or the same cost savings.

According to the Austin Business Journal, about a week ago, the University of Texas took the wraps off of "Lonestar," the "most powerful supercomputer for academic research"-- well, in Texas, anyway. It's nowhere near as big as Virginia Tech's "Big Mac," since UT's cluster consists of only "300 computer servers from Round Rock-based Dell Inc." A press release from July notes that each server houses dual Xeon processors (at an unspecified clock speed), so it's a 600-processor cluster, versus the 2,200-processor cluster at Virginia Tech. Gee, and we thought they built everything bigger in Texas; clearly we were misinformed.

There's no mention of the cluster's LINPACK score-- we'll have to wait for November's standings to know for sure-- but with a theoretical peak performance of only 3.67 teraflops, Lonestar would probably only rank somewhere between 18th and 30th on last June's list of the top 500 supercomputers. Theoretically, Virginia Tech's cluster is 4.8 times more powerful, and even when Lonestar bumps up to 1,000 processors by the end of 2004, it'll still lag by nearly a factor of 3. And here's the kicker: Virginia Tech spent $5.2 million on Big Mac; guess how much the Dell-based Lonestar is going to cost?

$38 million.

Yup, Lonestar cost over seven times as much as Big Mac, and theoretically delivers less than a quarter of its performance. So much for Dell always winning on price, hmm?

You know that age-old IT maxim that "no one ever got fired for buying Intel"? That may change now that buying Intel can represent a loss of millions of dollars. By the way, Japan's Earth Simulator, which currently ranks as the fastest supercomputer on the planet, is theoretically 2.3 times as powerful as Big Mac-- but it reportedly cost about $350 million to build. If G5s had been around last year, the Japanese could've built three Big Macs instead for about $15.6 million. Chronological disparity (and the fact that the Earth Simulator doesn't actually use Intel chips) aside, do you suppose whoever greenlit the budget on that project is going to get fired for overspending by over a third of a billion dollars? Maybe not, but we bet an end-of-the-year bonus is out of the question...

Hey, are you ready for a dramatic deflection into profoundly dissimilar subject matter? After yammering on about Mac-based supercomputer clusters for two entire scenes, we're going to shift gears into the entirely different and wholly unrelated subject of Mac-based parallel and distributed computing-- so buckle up, because the transition is going to be really jarring! If you don't strap yourself down to something heavy, you're sure to be thrown clear!

Oh, wait... We just noticed that Mac-based supercomputer clusters and Mac-based parallel and distributed computing are actually almost completely the exact same thing. The only difference is that while you're probably not likely to blow $5 million on an 1,100-node supercomputer for your basement, if you're a serious geek (in the nicest possible sense, of course) you might want to parcel out a single big rendering job or something over a half-dozen idle Macs in your lab-- the principle's the same, though. So never mind about that "strap yourself down" thing (unless you were planning to do it anyway). Gee, apparently here at AtAT, it's not just Columbus Day-- it's also All-Distributed Computing Day! (What, no iCards?)

So here's the skinny: faithful viewer David Triska pointed out that Think Secret has a quick blurb on something called Xgrid, which is apparently going to be an Apple tool for-- all together, now-- "parallel and distributed high performance computing." No, Apple hasn't announced any such product yet, but before your inner skeptic dares to pooh-pooh the notion, we should probably mention that Think Secret's source is pretty compelling: it's Apple's own public web site. The company posted a sign-up form for a new mailing list called "xgrid-users," which is, stunningly enough, for "discussions on using Xgrid." Go figure.

There's no telling whether this is another web lackey slip-up like the Great Power Mac G5 Specs Leak or just a low-key-and-no-big-deal intentional announcement on Apple's part, but we suspect the latter, because at broadcast time, the form was still up. (Given that Xgrid clearly isn't targeted at the average Mac user, it isn't exactly Expo Stevenote "One More Thing" fare, so there isn't much point in keeping it a crazy-big secret.) At this point, any details about Xgrid are wild speculation, but the name implies that it might plug into Panther's Xcode developer tools to allow tinkerers to build their own distributed applications. Or maybe it'll be optimized for flocks of Xserves so companies like Pixar can pick up a few truckloads of cluster nodes and get a render farm up and running in no time flat.

Who knows? We're just glad that the list sign-up page mentions the focus on parallel and distributed computing, because otherwise we'd have assumed that Xgrid was a G5-optimized port of Minesweeper. Or maybe Tic-Tac-Toe.