   |
|||
|
|||
|
:
News
: Archives
|
|
GeForce 8800 Technology Preview |
|
|
Massively General |
|||
|
|
|||
|
by Josh Walrath |
|||
|
Not much takes me by surprise anymore, but this launch certainly caught my attention. We have seen a fairly linear progression of technology in the graphics world, and few things have really ever shaken it up. The first 3D accelerators were simple, single pipeline, fixed function devices that were able to color and texture a pixel, and provide some bilinear filtering to a scene. As time went on we saw multi-texturing implemented, then twin pixel pipelines, then quad, hardware T&L, programmable integer pipelines, then onto programmable floating point pipelines. During that time we have seen designs become “wider” and faster, all the while offering more and more features. But there were few changes that could be considered radical. The previous generation is represented by the very impressive X1x00 and G7x series of chips from ATI and NVIDIA respectively. While both had distinct features over the other, they were still very similar in their overall layout and functionality. The R580 featured 16 ROPS, 16 texture units, 48 pixel shader units, and 8 vertex shader units. The G71 also had 16 ROPS, but differed from the R580 by only having 24 pixel pipelines with the ability to provide 24 texture units. When it came to performance the two architectures were surprisingly well matched. While the R580 had a more advanced feature set, the G71 more than held its own. These features aside, the two architectures still had the same basic pathway for pixel generation.
The GeForce 7 series all embraced a more traditional, but still largely parallel, rendering process. With the upcoming G80, this diagram is going to change drastically. The basic progression of pixel rendering was that the geometry was determined by the CPU, sent to the GPU for vertex operations, triangle setup, pixel shading and texturing, then off to the ROP for blending/anti-aliasing/z-buffering, and finally the frame was outputted to whatever device was connected to the video card. The architectures from ATI and NVIDIA, though with some distinct differences through the years, followed this same general course. We are now entering a new series of products that radically change the progression of pixel generation. I firmly believe we are seeing a paradigm shift in the graphics industry that is comparable to the first generation of consumer 3D graphics cards. These new products are an inflection point in graphics, as these products will not only be able to produce breathtaking images, but provide new and exciting functionality well apart from the traditional graphics pipeline. So why am I so excited about this technology? Let us explore the reasons…
Next: Architectural Expectations If you have found this article interesting or a great help, please donate to this site.
Copyright 1999-2006 PenStar Systems, LLC. |
|||
