Come on NVIDIA why is it so hard for you to............

[IBMer]

I'm glad to hear it does. It probably means it will run well on my MX when I get it.

That is ok, because the V5 plays better in D3D than it does in Glide.


Pardon me, but I always thought Sharky weren't the only review site around.

Sorry the Accuracy test wasn't a benchmark.

In the test the V4/V5 and the G450 both proved to be more accurate rendings than the reference. The Geforce was closer to the reference than the all the other cards. To bad the reference was flawed.

That's funny. Two paragraphs ago you were telling me all you care about is Sharky's review and you denouced the others. Now when it's convenient you ignore Sharky and go somewhere else to back up your arguments. So which is it? Are we using other reviews or sticking just to Sharky?

I am sorry, I pointed out that Sharky made a mistake. A mistake that will be correct as, I have e-mailed them about their mistake.

Your dodging the issue that you were FLAT OUT wrong.

The Voodoo3 box says: 32 bit texture rendering pipeline, bumpmapping and tri-linear mip-mapping. Yet in real tests it fails every one of them. Have you ever heard of marketing BS? Because 3dfx are really good at using it.

The Voodoo 3 can do Emboss bumpmapping, can internally render textures in 32-bit(if you don't think there is a difference try turning on 32-bit textures in Q3 and compare to the 16-bit ones. There is a big difference. It can also do Mip Map dithering, which is the same thing that the Trilinear filtering is on the GTS until you turn on Anisotropic filtering.

I think T&L is good so why haven't 3dfx used it? That would make 3dfx inferior, according to your own logic.

Are you an idiot.

First you say that the GTS has EMBM. Then you say because I say it doesn't that I am saying that 3dfx is inferior. Just look how that makes you look.

It supports them a hell of a lot better than the V4/V5 does.

When you can run the X-Isle demo on the V5 then you can speak. Also a lot of people have faster processors today. I think it would run fine on a 1.2Ghz T-bird which cost LESS than a GTS Ultra.

It's called memory bandwidth limits. You should know because the V5 suffers more of it than the GF2 does becasue of texture duplication and multi-processing overhead.

I am sorry, if it is memory limited or not then it should truely read what it is able to do, not what it could do if it had 1Ghz DDR Memory. BTW, the V5 has twice the bandwidth BECAUSE of the duplication of textures. In 3dMark2000 it is also MUCH closer to the fillrate on the box in both Single and Multitexturing applications than the GTS. How does it mean the Memory is much worse for the V5. Again, you do not provide proof for what you speak. Again, you speak in FUD.

You know what I think you are? I think you don't have a clue what you are talking about. You just tag along with the guys at Beyond3D and then come in here and regurgitate their arguments. The problem with this is that you take their arguments way out of context and you just come off looking like an idiot. So do us a favour and stop pretending to be an expert.

I don't think you have a clue of what your talking about. You pretend to know things and don't even have site, review, whitepapers or personal experience to back any of your claims up. I have provens some of your points to be COMPLETELY false and you tell me I have no clue.

 

[Dulanic]

IBMer do some research! 3DMark 2000 means crap in Win2K! I get the same FPS in Deus Ex in D3D in 98 and Win2K.... I count real life performance NOT a stupid benchmark.

 

[IBMer]

IBMer do some research! 3DMark 2000 means crap in Win2K! I get the same FPS in Deus Ex in D3D in 98 and Win2K.... I count real life performance NOT a stupid benchmark.

Why do I need to do some research when I have a GTS right here, and I get HALF the performance in D3D, I play in OpenGL though.

 

[BenSkywalker]

IBMer

"Actually your incorrect.

8-tap Anosotropic filtering is TRUE trileaner filtering.

NVidia uses Mip Map Dithering to simulate tri-linear filering, but only 8 sample is true trilinear filtering."

Anisotropic is different then trilinear, where are you getting your information from? I looked for some simplified terms on sites(You can check here for an in depth look on anisotropic) and found these-

Guru3D

"Trilinear filtering will combine Bilinear filtering in 2 Mip levels. This however results in 8 texels being needed so memory bandwidth is multiplied by 2. This usually means that the memory will suffer serious bandwidth problems so trilinear filtering is usually used as an option."

"Anistropic Filtering addresses quadrilateral shaped and angled areas of a texture image. A sharper image is accomplished by interpolating and filtering multiple samples from one or more MIP-maps to better approximate very distorted textures. This is the next level of filtering after trilinear filtering. While it will create the best looking images it comes at a serious price and should only be used when your system can handle it. If your system is performing slowly try turning on and off Anistropic filtering for better perfomance."

Trilinear could be deemed to be "dithering of mip maps", you could argue that any mip mapping at all is dithering of textures for that matter.

There is a difference between trilinear and anisotropic, and nVidia does do them both(as does ATi). Where did you get the information that nVidia is not handling both properly or that they(trilinear and anisotropis) are the same?

 

[Dulanic]

Well then you must #1 be using older drivers #2 Saying your half performance is in 3DMark and not games. 3DMark doesnt work right in Win2K.