I posted this over at RAge3D first (oddly enough, that was Hellbinder thread too) but I feel it is appropriate here as well.
This is something that has been bothering my for some time & I'm glad that you started this thread, Hellbinder, my remarks are somewhat OT but I feel that they are tangiant enough to go into here.
What I am am obtusely referring to, of course, is process technology. Although I was a double E major in college (Ohio University...GO BOBCATS!!!) I flunked out of Calc II & thus today I have no actual expertise in this field. However, I have seen many skeptics her & at Beyond3d (particularly some whose names start with CH) stating that the NV30 must be better than the R300 because it has a)X more transistors & the biggie, b)the .13 process. I would say to all you EE doctoral candidates, you should get your money back. Apparently ATI has managed to squeeze 400MHz out of that oh so inefficient .15 process. Many naysayers were saying 325 would be impossible.
As an AMD (& AYTY) investor I clearly remember the clamor that the PIV must be better because it was so much bigger & had to many more transistors & it was so much faster. Well we all know how that scenario played out don't we.
How do any of you know that the NV30 is better, faster, the proverbial 6 million dollar chipset. As an investor, one of the cardinal rules is past performance is no indicator of future results!.
There are alot of blanket assumptions being made & so far from nvidia we have what!! Poof, nothing, nada. As a matter of fact, I would contend from the current things that we do know, the .13 process has proved exceedingly difficult (Intel was set back approx 9 months, amd almost 12 & there is no way that any of the fab-for-hire guys are any better than those two), yeilds are atrocious, & as of yet, there is only empircal data that .13 is better than .15. Furthermore I would contend that the real reason behind the lateness with the NV30 is that ATI caught nvidia flatfooted (much like AMD did to Intel when the uncrippled Athlon came out), & that they are now in the midst of a minor, but not unsignificant design change.
The point of this diatribe is that today, ATI has a verifiable, reproducable results from a product that has surpassed the expectations of the most ardent of fanATIcs. Could not the NV30 go in the other direction?
