Yuku free message boards

Quote:

---

I personally will stick with the 9700pro. Sorry Nvidia, you lost your crown to Canada's ATI after you decimated 3dfx. The 9700 was meant to compete with the Ti4600 not the FX but it still steps all over the FX. Wait till ATI release its next video card shortly. Nvidia you are toastidos!

---

Actually, no... the 9700Pro was never intended to compete with the GeForce4 Ti 4600, as the GeForce 4 series (bar the MXs) are DirectX 8.1 while the Radeon9xxx series (bar 9000-9100) is completely DirectX 9. Thus, results from benchmarks like 3DMark03 are not comparable since the GeForce4 cannot do all the tests.

Futhermore, NVIDIA isn't "toastidos" as you have stated. They own a huge portion of the OEM market which generates a lot of revenue, as you might imagine. The GeForce4 series will continue to sell, considering that the prices have been slashed quite a bit since the GeForceFX debuted. On top of that, they have generous amounts of money coming in from their nForce2 AMD motherboard chipsets. The GeForceFX may be a setback, but it is definately not the last of NVIDIA.

Let's not forget that ATi has yet to jump hurdle known as the 0.13 micron process. They can keep bumping up that 0.15m chip of the Radeon9500-9700, but eventually they're going to hit the point where they can't "bump it up" anymore. We'll see how hard it is for them to produce a 0.13m chip, won't we?

Looks like I wasn't entirely correct. nVIDIA officially got caught cheating in the 3DMark03 benchmark (and unofficially in the 3DMark2001 benchmark).

Conterpoint

#1 - They do know what they are doing, but the fact of the matter is that they thought no one else would know what they are doing, which is cheating

#2 - That's a valid point, but no one ever investigated the idea of nVIDIA cheating in their drivers... especially when there was no other video card to challenge it's power (referring to the release of the 23.11 drivers with GeForce3). nVIDIA could have been replacing shaders in 3DMark2001 with their own "optimized" shaders, but no one bother to investigate further.

Furthermore, Futuremark now has hard evidence that nVIDIA was cheating in 3DMark03. Using back buffer clears, clipping planes, and optimized shaders in NOT allowed when running a benchmark. Every peice of hardware that runs the benchmark should do the exact same amount of work to provide an accurate comparison between the hardware. When a company adds in code that tells the hardware to do less work on the benchmark, it does not provide an accurate comparison between the hardware. Moreover, usually adding "optimizied code" leads to image quality degradation and artifacts. What even more is that the latest drivers do not even render some of the images that a user would normally see in the benchmark! For example, in GT2 of 3DMark2001, the dragon is completely unrendered which allows nVIDIA to claim drastic improvements in framerate.

#3 - I'm sure nVIDIA has a totally seperate department for the video card design and motherboard design.

As far as I know, game makers do not sign contracts with video card companies so a certain video card will achieve better performance. Game makers test games on a variety of hardware so that all consumers are able to play the game, not just ones with specific hardware. Let's say a game maker made a deal with Intel so that the game will only run on Pentium CPUs. The game maker would lose a huge amount of profit because so many people use AMD processors. Now take that example and insert nVIDIA for Intel. The effect is still the same.

Futher, nVIDIA would actually be losing profit by doing this because they would have to pay the game maker to insert specialized code for nVIDIA hardware. How much do you think it costed nVIDIA to put the "nVIDIA: The Way It's Meant To Be Played" logo in UT2003?

Finally, your "next three years" theory is counter-pointed by what I have previously said. Game developers are not going to waste huge sums of profit to design on one specific peice of hardware. They design games according to their own specific plan so they can obtain the best amount of compatibility for a variety of computer setups. That way, more consumers will be able to play their games and the company will make more money.

nVIDIA is actually playing defense now. They've no arguement for the reason why their hardware is losing so much performance in many common benchmarks (see this article: discuss.futuremark.com/forum/showflat.pl?Cat=&Board=techdisplayadapters&Number=2404382&page=0&view=collapsed&sb=5&o=0&fpart=1) that they have resorted to saying any benchmark they lose in is not an accurate tool for measuring their hardware.

Quote:

---

the contracts are not to make the games only run on geforce rigs but to be optimized on geforce rigs. Game Makers will loose no revenue on this move.

---

Why would they do this? nVIDIA would pay huge sums of money to the developer to do this, and it would slow down the entire game making process altogether.

Wouldn't it make more sense for a developer to produce a game with no optimizations (or optimizations for every peice of hardware) so more consumers would be inclined to buy their products?

On another note, I thought that the GeForceFX 5900 was a great card until all the stuff nVIDIA has done to invalidly boost its scores. Now I'm not really sure the power of this thing because of all the cheats nVIDIA has pulled recently.