Well, after yesterday's spiel about vague and sketchy PowerPC 970 rumors, the equal time laws require that today we point you toward something a little more substantial. Isn't it a stroke of luck, then, that faithful viewer Joseph Rosmann just tipped us off to the latest compendium of solid, reasoned 970 analysis over at Ars Technica? The author notes that it's actually longer than his master's thesis, and this is only the second half. Zoinks. In addition, it's apparently all based on-- get this-- actual fact. Go figure.

Yes, somebody who knows a heckuva lot about chip design and performance pored through all available public data on the 970 and spent a few months working up this vast and detailed document describing the chip in excruciating detail, plus providing some (extremely well-) educated guesses about how it'll rate compared to the existing G4 and the Pentium 4. For us, however, there's just one little problem: long and detailed technical documents about processor design fall into a category of reading material we like to call "Verbal NyQuil." They may not alleviate our sniffling, sneezing, coughing, aching, stuffy head, or fever (more's the pity), but they sure can turn out our lights faster than a sock full of quarters to the base of the skull.

Luckily, the author understands that some of us poor uneducated slobs with attention spans measurable only in nanoseconds might still be interested in the big picture, and he was kind enough to post links to the good bits. As far as we could make out (in between some quality naps), any Mac fans expecting the 970 to wallop Intel's best into a bloody smear are likely to be disappointed-- but it should definitely make great strides in narrowing the Apple-Wintel performance gulf back into a mere performance gap. (Fear not, though; the 970 ought to provide Apple's marketing gurus with plenty of ways in which to make the Mac look faster. Maybe even much faster. Let the Photoshop bakeoffs begin!)

For us, though, the best bit in the entire article is the author's evidence that Apple will actually even use the 970 in the first place. Remember, neither Apple nor IBM have officially said that there'll ever be 970-based Macs at all-- we're all just assuming that Apple might want Macs to be fast again or something. But the fact that the Altivec unit was apparently grafted onto IBM's original POWER4 core like some kind of freakish second head does indeed hint that this chip was actually designed to Apple's order and destined to be used in Macs. Of course, now that we're feeling all smiley about the "if," we're antsier than ever about the "when." Hey, Apple-- are we there yet? Are we there yet? Are we there yet?...