Looking Back Pt. 2: X800 & Catalyst Under The Knife
by Ryan Smith on February 22, 2006 12:05 AM EST- Posted in
- GPUs
Doom 3
Moving back in to first-person shooters, Doom 3 is without a doubt the most interesting game tested on the Catalyst drivers today. With its emphasis on darkness and a unified lighting system, Doom 3 presents a very different situation than most first-person shooters do, hopefully giving us a different take on performance in such a game. It’s also the OpenGL game of choice for this roundup; though, as we’ll see, just being OpenGL doesn’t mean it’s a great indicator of OpenGL performance.
Shortly after Doom 3 was released, ATI found itself in an interesting situation with regards to what they could do to improve performance. There was a bottleneck in the game in how specular highlighting was applied, and while ATI made some efforts to optimize their drivers for the game, as seen with the 4.09 drivers, this kind of bottleneck was a fundamental issue on which ATI would have to take more serious measures if they wanted to remove it.
What ended up being the bottleneck was that John Carmack, id’s lead programmer, had decided to use a lookup method for determining what highlighting values should be used, based on referencing a specifically constructed texture map with these pre-computed values. It turned out that the R420 could actually calculate such values faster than it could look them up, so to fix the bottleneck would mean replacing the entire shader with what was only a mathematical approximation for the real values in the texture map. However, given the scrutiny over optimizations, it is a difficult choice to make. We’ve covered the issue before, but ultimately, these optimizations are valid in most cases, and the result is the performance improvement as seen above. This case, however, will always serve as a reminder of how fine of a line there is between optimizing and cheating in a game.
But getting back to performance as a whole, outside of ATI’s shader replacement, there’s no further changes in performance. Unfortunately, the replacement means that Doom3 isn’t too great of an OpenGL benchmark, but as the number of OpenGL games continues to dwindle, there is little else on the market to play that uses OpenGL, which isn’t Doom 3 (or Doom 3 engine based) in the first place.
Moving back in to first-person shooters, Doom 3 is without a doubt the most interesting game tested on the Catalyst drivers today. With its emphasis on darkness and a unified lighting system, Doom 3 presents a very different situation than most first-person shooters do, hopefully giving us a different take on performance in such a game. It’s also the OpenGL game of choice for this roundup; though, as we’ll see, just being OpenGL doesn’t mean it’s a great indicator of OpenGL performance.
Shortly after Doom 3 was released, ATI found itself in an interesting situation with regards to what they could do to improve performance. There was a bottleneck in the game in how specular highlighting was applied, and while ATI made some efforts to optimize their drivers for the game, as seen with the 4.09 drivers, this kind of bottleneck was a fundamental issue on which ATI would have to take more serious measures if they wanted to remove it.
What ended up being the bottleneck was that John Carmack, id’s lead programmer, had decided to use a lookup method for determining what highlighting values should be used, based on referencing a specifically constructed texture map with these pre-computed values. It turned out that the R420 could actually calculate such values faster than it could look them up, so to fix the bottleneck would mean replacing the entire shader with what was only a mathematical approximation for the real values in the texture map. However, given the scrutiny over optimizations, it is a difficult choice to make. We’ve covered the issue before, but ultimately, these optimizations are valid in most cases, and the result is the performance improvement as seen above. This case, however, will always serve as a reminder of how fine of a line there is between optimizing and cheating in a game.
But getting back to performance as a whole, outside of ATI’s shader replacement, there’s no further changes in performance. Unfortunately, the replacement means that Doom3 isn’t too great of an OpenGL benchmark, but as the number of OpenGL games continues to dwindle, there is little else on the market to play that uses OpenGL, which isn’t Doom 3 (or Doom 3 engine based) in the first place.
Catalyst 4.05 versus 6.01 (mouse over to see 4.05)
Facebook
Twitter
RSS
24 Comments
View All Comments
lombric - Thursday, February 23, 2006 - link
It may be interesting to see the evolution in cpu discharge under various video format and in image quality.I know that the introduction of AVIVO in recents drivers was very efficient for the X1xx serie but what about the R420? No chance to have similar results?
Egglick - Friday, February 24, 2006 - link
As far as I know, the X1x00 cards are the only ones with AVIVO, or at least the entire feature set.So does that mean that a $80 X1300 has better video playback than a X850XT PE?? Yep.
pkw111 - Thursday, February 23, 2006 - link
... but their conclusion is rather boring. True it may be good solid research, but how about some studies that give colorful results, liek comparing the non-offical ATI drivers, such as WarCat, Omega, ngo, etc.Egglick - Thursday, February 23, 2006 - link
I think it's a little too early to make guesses about the R5xx series right now. Don't forget that both the X800Pro and the 9700Pro are R300 based, and what we're looking at is a cumulation of 3+ years of tweaking and optimizing. The R580 has been out for what, a month?We could still see very radical performance boosts for R5xx based cards, particularly the R580 with it's unique shader architecture. It's also possible that performance boosts in new games will be even larger once the successive driver has been optimized for it. Basically, it's a whole new architecture, and what may have been true for both of these R300 based cards may not be true at all for R5xx.
Also, the CCC is garbage. Boo to ATI for forcing us to use it.
DieBoer - Thursday, February 23, 2006 - link
I just wish ati would stop wasting time on optimising 3dmark and start with games. No serious gamer would take notice at all at 3dm scores only the average joe.Spoelie - Thursday, February 23, 2006 - link
The most horrifying thing about CCC is the horrendous memory usage. I had been using the normal control panel all this time but recently formatted and downloaded the latest drivers. WindowsXP's memory usage after bootup went from ~70 something (not much had been installed yet) to a full fledged 200mb!! Only from installing the f*cked up driver.After some tweaking (disabling all ATi's added services and the CCC entry in the registry's startup) I'm back at around ~95mb after startup, which I was at before the format.
Still find it incredible in what kind of default configuration the CCC 'ships'.
Questar - Thursday, February 23, 2006 - link
Your saw rhings that weren't there - XP's footprint is much larger than 70MB.Spoelie - Friday, February 24, 2006 - link
Not really, once you start tweaking and don't have all programs installed, around 70 is really not that much of a stretch without programs open. Even so, even if the task manager for some reason is lying about the absolute numbers, there was a difference of 130mb just by installing a driver.abhaxus - Wednesday, February 22, 2006 - link
i find it surprising that you did not run the test with a dual core CPU to see if the dual core optimizations actually did anything in the new drivers. i know there was a writeup on them awhile back with the 5.12s i believe but i'd like to see if newer versions got any further improvement.SonicIce - Wednesday, February 22, 2006 - link
You can create a short 300 frame timedemo for Farcry and play it back with the http://www.hardwareoc.hu/index.php/p/download/st/....">Farcry bench tool in screenshot mode. This will give you perfectly consistant results. I did it once to compare the shadows on the weapon.