Heb alles maar gepast omdat her en der alle info weer even snel verdwenen is als dat het gekomen is.
De info is afkomstig van HardOCP, Apple en VR-zone.
Elsa NV20
De info is afkomstig van HardOCP, Apple en VR-zone.
<< NVIDIA is also using a form of Hidden Surface Removal, or what you may know as "Tiling", but the GeForce3 does not use the current technologies mainly utilizied. Unless you have a Vid Card capable of Tiling or HSF, it is most likely the case that when your card is used for 3D applications, it is drawing a lot of things that you will never see on your monitor. Why don't you see this? Are we missing out? No, we are simply seeing something in FRONT of the object. Let's say you are playing your favorite 3D shooter and you are lining up a rail shot when another player walks into your line of sight. You can no longer see your target, but you and I know he is still there. Well, even though we can't see him, he is still being "drawn" by your Vid Card. The GeForce3 will no longer draw MOST of the stuff you don't see. Game programmers can specifically write their code to better take advantage of this feature to help out their game's performance. Does the GeForce3 do this 100% with any game? No, it does not, as there are times when the GPU is making a decision on whether an image will show up correctly on the monitor. As this gets programmed in by more developers, we will certainly reap the rewards.
Data compression is exactly what it says. The GeForce3's compression being LOSSLESS is worth mentioning, as compression usually means some sort of compromised image quality. The compression is done in hardware and only compresses for the trip between the GPU and DDR Ram. This frees up some of our pipeline so that the other three memory controllers get a better chance to kick in. Also, these technologies implemented above are invisible to ANY API, so we will see gains in both Direct3D and OpenGL titles. >>
<< NVIDIA will no longer be "blowing up" your 800x600 screen to 1600x1200 then shrinking it back down in hardware to do their antialiasing in their new Quincunx mode. I am not going to get into the specifics here, but from what I am hear, this is the mode you will be using. To give you a basic rundown, it gives you 4X AA quality at a 2X price. Now if you compare Quincunx mode with true 4X AA side by side, you will see a difference. The edges of images using Quincunx mode are a bit softer compared with the harder edges of the 4X mode. If you are wondering what a Quincunx is... quin·cunx (kwin'kungks) It is an arrangement of five objects, with one at each corner of a rectangle or square and one at the center. This is a reference to the way they sample the pixels that are used to fill in the jaggies. The GeForce3 pulls information from the pixels surrounding a pixel, in order to smooth out the lines that are not perfectly vertical or horizontal. I am sure you will be seeing some in-depth discussions on how this is done. All of this is also done in hardware. Gotta do something with those 57 million transistors. >>
<< Finally, you see NVIDIA referencing 70 frames per second using this Quincunx mode. Just for kicks, I took a couple of benchmarks to compare with a GeForce2 Ultra card. 1.35GHz AMD CPU and 128megs of Ram, all running on a 150MHz bus. With NO AA turned on, we got 130 fps. With 2X Sampling, we got 33 fps running Demo001. We will see if their "71 fps" comes close to reflecting the truth or not. >>
Elsa NV20
The nVidia GeForce3 graphics chip is the personal computer industry's most advanced graphics processing unit ever. With more than 57 million transistors it's able to perform 800 million operations per second including 76 billion floating point operations per second (GigaFLOPS). Powered by the new nVidia nfiniteFX engine and the LightSpeed Memory Architecture, the GeForce3 GPU enables the user to experience a much richer 3D graphics environment than they ever have before. Featuring a large 64MB DDR SDRAM frame buffer it can handle a new generation of games and professional applications that use large texture maps. Naturally the GeForce3 supports very high resolutions with full color support (up to 2048 x 1536 at 32 bits per pixel).
The high-end chip will cost US$600, will be available in late March, and can be ordered by customers or resellers, Apple CEO Steve Jobs said during his Feb. 22 Macworld Tokyo keynote. He said the GeForce 3 was the "most advanced graphic chip ever" (boasting 57 million transistors and a workload of 76 gigaflops).
:strip_icc():strip_exif()/u/7205/duke3.jpg?f=community)
:strip_icc():strip_exif()/u/182/crop66d4984812e9a_cropped.jpg?f=community)
:strip_icc():strip_exif()/u/4400/haai.jpg?f=community)
