ATI All-in-Wonder Radeon 32MB DDR
by Matthew Witheiler on August 21, 2000 3:00 AM EST- Posted in
- GPUs
The Core
As with every video card, something must power not only the TV side of things but also the computer side. In the past, the power of ATI's All-in-Wonder cards have been dwarfed in comparison to other video card offerings. Due to extra development time as well as frequent product delays, previous All-in-Wonder cards proved not to be very attractive to power users. In fact, the performance of the All-in-Wonder based products have historically lagged so far behind other video cards that some users actually found themselves buying not only the All-in-Wonder video card but also another manufacturer's card for use in games. Thankfully, with the current generation of All-in-Wonder cards ATI has completely fixed this problem How did they do that? By taking none other than their recently released Radeon core and making it the heart of the new All-in-Wonder Radeon.
Although we took an in-depth look at the Radeon core not too long ago in our ATI Radeon 64MB DDR review, let's refresh our memory as to what this powerful new core can do.
Three Texture Units
The Radeon core possess two rendering pipelines with three texture units per pipeline, resulting in a total of 6 texels per clock. This places the core 2 texels less than the quad pipeline, dual textured GeForce2 core that powers the GeForce2 MX, the GeForce2 GTS and the GeForce2 Ultra. Although now the Radeon core is actually crippled by the tri textured unit, due to the fact that the vast majority of current games use only two textures and leave the third texture unused, ATI is betting that future games will move to three textures, a transition that could leave the Radeon performing very fast and the GeForce2 in the dust. It looks like we might get to see if this is in fact the case with Microsoft's upcoming game entitled Halo, which supposedly uses three textures per pixel.
T&L
Similar to the GeForce2 core, the Radeon core possesses a T&L engine that is capable of performing transform, lighting, and clipping calculations right on the chip. In addition, the Charisma engine, as ATI likes to call it's T&L features, is able to perform both key frame interpolation as well as vertex skinning. Both are features that attempt to enhance the visual features of gameplay, it may be awhile before we see these features in action.
HyperZ
Perhaps one of the most advanced, as well as embraced features of the Radeon is its HyperZ technology. Aimed at reducing the amount of information that must travel over the video card's memory pipeline, HyperZ technology actually consists of three parts. First is Hierarchical Z, which attempts to reduce the amount of overdraw that is encountered in any given scene. Next HyperZ incorporates Z Compression, which losslessly compresses and decompresses the information stored in the Z-buffer, reducing the amount of data that is transferred during Z-buffer reads and writes. The final piece of HyperZ is Fast Z Clear, a feature that reduces the time necessary to clear the Z-buffer, a step necessary to render each new frame.
Look for an in-depth article describing HyperZ and its benefits in coming weeks.
Now that we have refreshed our memory about some of the technology behind the Radeon core, lets take a look and see how this core is used in the All-in-Wonder Radeon.
0 Comments
View All Comments