ATI Radeon HD 2900 XT: Calling a Spade a Spade
by Derek Wilson on May 14, 2007 12:04 PM EST- Posted in
- GPUs
Tessellation and the Future
It's no secret that R600 is AMD's second generation unified shader architecture. The Xbox 360 houses their first attempt at a unified architecture, and the R600 evolved from this. It isn't surprising to learn that some of the non-traditional hardware from the Xenos (the Xbox 360 GPU) found its way into R600.
AMD has included a tessellator on their hardware, which is able to take input geometry and amplify it before passing it on to the vertex shader. This is something that Microsoft is planning on adding to future versions of DirectX as well, but in the meantime developers will need to take special steps to utilize the hardware.
The basic idea behind tessellation is in the subdivision of geometry. There are multiple algorithms for handling this process, and the R600 tessellator is capable of adapting to a developer's specific needs. The tessellator can take a polygon as input and break it up into smaller triangles, creating more vertices for a specific object. Taken on its own, this isn't particularly useful, but this concept can be paired with displacement mapping in order to reshape the tessellated polygon into something more like the actual surface a developer wants to represent (this is called the limit surface).
With low polygon models and lots of pixel shaders, normal maps and textures can approximate the look of more complex geometry, but we're always stuck with the very rough silhouette edge around the object. With more geometry, we could also better use pixel shaders to enhance the geometry present rather than trying to create the illusion of geometry itself.
We can't simply send millions of polygons per character to the graphics card. This isn't because the card can't handle the processing requirements, but rather the bandwidth and latency overhead of sending all this data to the hardware is too high. Tessellation and displacement gives us a way of really using the vertex shading power of unified architectures as well as removing the limitation on polygon count created by overhead.
While geometry shaders can be used for amplification and tessellators can be written as geometry shaders, this process is still way too slow on current programmable hardware. AMD's dedicated tessellator is capable of tessellating up to 15x more data and it can work much faster and more efficiently than a geometry shader set to the same task. With the next version of DX bringing tessellator hardware to all GPUs, developers should be able to focus on more interesting uses for the geometry shader as well.
Having this unit makes porting Xbox 360 games even easier for developers targeting AMD hardware. As most hardware still doesn't support the feature, a more general purpose path will still have to be written, but there wouldn't be any reason to remove what's already there. In these cases, R600 could benefit with greater performance than other hardware.
The downside is that it might be difficult to entice developers not already working with the Xbox 360 to touch the tessellator. It is definitely capable of high performance and terrific detail, but spending time on a feature only a small subset of gamers will be able to experience (for this generation) takes away from time spent making the game better for everyone.
We are always happy to see either hardware or software take a leap and create the first chicken or egg, but we just don't see the tessellator as a big selling point of R600. The technology is great, we're glad it's there, but we will really have to wait and see just how much (if any) real value this adds to the product. We'll leave this section on one final note about a tessellator landscape demo that really brings home what this thing can do.
Facebook
Twitter
RSS
86 Comments
View All Comments
mostlyprudent - Monday, May 14, 2007 - link
Frankly, neither the NVIDIA nor the AMD part at this price point is all that impressive an upgrade from the prior generations. We keep hearing that we will have to wait for DX10 titles to know the real performance of these cards, but I suspect that by the time DX10 titles are on the shelves we will have at least product line refreshes by both companies. Does anyone else feel like the graphics card industry is jerking our chains?johnsonx - Monday, May 14, 2007 - link
It seems pretty obvious that AMD needs a Radeon HD2900Pro to fill in the gap between the 2900XT and 2600XT. Use R600 silicon, give it 256Mb RAM with a 256-bit memory bus. Lower the clocks 15% so that power consumption will be lower, and so that chips that don't bin at full XT speeds can be used. Price at $250-$300. It would own the upper-midrange segment over the 8600GTS, and eat into the 8800GTS 320's lunch as well.GlassHouse69 - Monday, May 14, 2007 - link
If I know this, and YOU know this.... wouldnt anandtech? I see money under the table or utter stupidity at work at anand. I mean, I know that the .01+ version does a lot better in benches as well as the higher res with aa/af on sometimes get BETTER framerates than lower res, no aa/af settings. This is a driver thing. If I know this, you know this, anand must. I would rather admit to being corrupt rather than that stupid.GlassHouse69 - Monday, May 14, 2007 - link
wrong section. dt is doing that today it seems to a few peoplexfiver - Monday, May 14, 2007 - link
Hi, thank you for a really in depth review. While reading other 'earlier' reviews I remember a site using Catalyst 8.38 and reported performance improvements upto 14% from 8.37. Look forward to Anandtech's view on this.xfiver - Monday, May 14, 2007 - link
My apologies it was VR zone and 8.36 to 8.37 (not 8.38)GlassHouse69 - Monday, May 14, 2007 - link
If I know this, and YOU know this.... wouldnt anandtech? I see money under the table or utter stupidity at work at anand. I mean, I know that the .01+ version does a lot better in benches as well as the higher res with aa/af on sometimes get BETTER framerates than lower res, no aa/af settings. This is a driver thing. If I know this, you know this, anand must. I would rather admit to being corrupt rather than that stupid.Gary Key - Tuesday, May 15, 2007 - link
I have worked extensively with four 8.37 releases and now the 8.38 release for the upcoming P35 release article. The 8.37.4.2 alpha driver had the top performance in SM3.0 heavy apps but was not very stable with numerous games, especially under Vista. The released 8.37.4.3 driver on AMD's website is the most stable driver to date and has decent performance but nothing near the alpha 8.37 or beta 8.38. The 8.38s offer great benchmark performance in the 3DMarks, several games, and a couple of DX10 benchmarks from AMD.
However, the 8.38s more or less broke CrossFire, OpenGL, and video acceleration in Vista depending upon the app and IQ is not always perfect. While there is a great deal of promise in their performance and we see the potential, they are still Beta drivers that have a long ways to go in certain areas before their final release date of 5/23 (internal target).
That said, would you rather see impressive results in 3DMarks or have someone tell you the truth about the development progress or lack of it with the drivers. As much as I would like to see this card's performance improve immediately, it is what it is at this time with the released drivers. AMD/ATI will improve the performance of the card with better drivers but until they are released our only choice is to go with what they sent. We said the same thing about NVIDIA's early driver issues with the G80 so there are not any fanboys or people taking money under the table around here. You can put all the lipstick on a pig you want, but in the end, you still have a pig. ;-)
Anand Lal Shimpi - Monday, May 14, 2007 - link
There's nothing sinister going on, ATI gave us 8.37 to test with and told us to use it. We got 8.38 today and are currently testing it for a follow-up.Take care,
Anand
GlassHouse69 - Monday, May 14, 2007 - link
wow dood. you replied!Yes, I have been wondering about the ethics of your group here for about a year now. I felt this sorta slick leaning towards and masking thing goign on. Nice to see there is not.
Thanks for the 1000's of articles and tests!
-Mr. Glass