Didn't have much time last nite to reply, now I can get to some points I wanted to cover.
BFG
My assupmption on your comments on 2D quality were the issues you were having with 640x480 16 color 60HZ, if that is not the case then no, that wouldn't be a mobo caused issue. Everything else you list I have seen on a GF DDR on one of my POS mobos, an old SS7 Soyo SY5EMA. Additional problems, not waking up from sleep mode, entire windows leaving massive screen corruption making the system unuseable, constant hard locks and crashes, icons refusing to do anything when clicked on, I could keep going.
None of these problems were had with my TNT, but were when I placed my GF DDR in. From there I moved to a mobo I knew had a POS AGP slot, the Asus K7M. Again, most of the problems you are having were also had by myself until I dropped it to AGP 1X. It was clearly the fault of my mobo and its' sh!tty AGP slot, not nVidia's. I ran the same board in a BH6 and it is currently operating in a K7TPro2A flawlessly. You need a clean power supply to the slot, and an adequate one, for any board that utilizes the AGP slot(instead of just plugging in to it). None of these issues were caused by nVidia, they were from sub par mobos(Soyo) or at the very least crap AGP slots(Asus).
I don't think 3dfx should be blamed for the PIV issues. It isn't any vid cards company duty to make sure that all mobos conform to industry standard specs(well, nVidia will be responsible for their own mobos).
Blackhawk2-
"And when isn't this the case?"
Do yourself a favor and click on the link to look at the game we were talking about. Then explain where the overdraw is. In most of the screenshots there is no overdraw for the majority of the screen, just a background.
"Typical overdraw these days is around 3 and in the future it will only get worse."
Says who? You don't think software developers are using their own techniques to deal with this? In the future, eight pass texturing and the likes is what makes boards more sensitive to overdraw, not that the actual amount will increase. It is possible, but to date nothing I have seen supports this notion.
"We'll see what the Kyro2 brings to the table."
Of course we will.
"Thats only partially correct when you consider 3dfx's Full Scene Anti-Aliasing. I wonder how 3dfx's T&L would have compared to Nvidia's T&L."
So they had one feature that everyone else does too.... what?? There implementation is a bit better then the competition. They lose out in trilinear and anisotropic just looking at the basics, they don't even offer it in any multitextured games. Having an edge in one feature negates being late with every other one for the last two and a half years??
"Nvidia is more focused on making money than making good products, in the end its going to bite them in the @ss."
Why don't you, since you know so much about 3D technology, talk to SGI, the founder of the overwhelming majority of principles we use for 3D right now, about how nVidia doesn't make quality products. Why don't you tell them that nVidia's products aren't good enough to use for a gaming card so they simply can't be good enough for professional 3D animation and CAD workstations. While your at it, you better call up Pixar and tell them to send back the large batch of workstations that they ordered to help them continue to push CGI forward using nVidia based hardware. Since you know so much, certainly they should all listen to you, right?
"A good example of Nvidia's haste to get a chip to market while creating a lesser quality product."
You will inform SGI and Pixar that they are using low quality hardware right? They seem to be under the impression that it is great, but what the he!! do they know about 3D?
"A good example of Nvidia's haste to get a chip to market while creating a lesser quality product."
Huh? WTF does transistor count have to do with the price of tea in China? The 486DX has a he!! of a lot less transistors then a Duron. Thunderbird or PIII/PIV so it must be a lot higher quality product, right? Your logic here.
"Sucking "a lot of juice" has the potential of causing heat problems. We'll have to see how Nvidia handles that but so far I've been losing faith in Nvidia's ability to see past their own forehead"
The NV20 uses LESS power then the V5 5500, as in, not as much. The NV20 produces LESS heat then the V5 5500. Who couldn't see where? As a matter of fact, the NV20 uses less power then the NV10 did, are you saying that nVidia should take the time and care to significantly increase their power useage and heat production as 3dfx did from the V3 to the V5?
Soccerman
"the game isn't THAT much better (which was why I was kind of questioning the polygon count).. in fact, what impressed me the most were the explosions (so you were right ."
The game looks very poor compared to Giants to me. If you want to see a good example of what PC gaming can look like right now, check out some reviews with screenshots(or the game itself if your rig can handle it). It makes Quake3/UT and the like look seriously dated.
"I'm not saying a Kyro 2 would really benefit from that kind of game, it's just, I want a Kyro 2 for ALL games..."
Unless the KyroII is in fact a new part that shares very little, if anything, with the original outside of the name, or is running much faster then 166MHZ it will not be anywhere close to the NV20.
"besides, games like that one aren't that Video card intensive, correct? I've always noticed that Space/Flight sim games have been pretty slow overall, (if you ask me) probably becuase of low CPU speed"
True, they are not video card intensive.
<rant mode> You know that they could easily be so if developers would exploit the fact that they were CPU limited. They could use extremely complex models by offloading them to the T&L unit(sticking with static vertices) and throw out every texturing trick in the book along with a barrage of every particle effect they can pull off as the CPU will still be dealing with game code and you won't be vid card limited. Your CPU is choking while your vid card is "twiddling its' thumbs".</rant mode>