Adapting to Parallelism: Catalyst 5.12 More Dual Core Friendly?
by Derek Wilson on December 4, 2005 10:45 AM EST- Posted in
- GPUs
Introduction
Friday afternoon we got an email pointing us to the latest beta driver from ATI. One of the key features of this driver is a performance boost for dual core systems. Building a driver to take advantage of parallel processing is quite a task, and extracting any noticeable performance gain out of it is even more difficult. So we are here today to see just what ATI has gotten out of their efforts thus far.
Admittedly, the highest performance gains come from low resolutions without antialiasing. It stands to reason that the more CPU limited a test is, the more benefit the game will get from freeing up CPU resources. The real benefit to end users if only lower resolutions benefit is questionable, but every step helps. With the future of computer hardware firmly planted in parallelism, the burden of improving performance shifts a little further towards software developers. Coming up with new and interesting ways to parallelize code efficiently is going to be quite a new task for desktop software programmers to tackle. And in the end, Amdahl's Law reminds us that we are still limited by the benefit we can get from parallelism. The percentage of code that must remain sequential will become the limiting factor. But every little bit of parallelization still helps.
There are really quite a few questions to be asked about this driver. After adding up everything we wanted to do, the sheer number of tests we had laid out was enormous. In an effort to be more efficient ourselves, we decided to break our analysis of the 5.12 driver up. This article is meant as a quick look at the benefits of ATI's dual core enhancements on a few select games running on X1K series hardware. We will compare this driver to the old one as well as dual core performance to single core performance.
Our next look at the 5.12 driver will include a comparison to NVIDIA performance in single and dual core systems, more than one ATI card, more games, and as many more things as we can pack in. Of course we are open to suggestion. But for now, we'll take a look at what we've got to work with.
Friday afternoon we got an email pointing us to the latest beta driver from ATI. One of the key features of this driver is a performance boost for dual core systems. Building a driver to take advantage of parallel processing is quite a task, and extracting any noticeable performance gain out of it is even more difficult. So we are here today to see just what ATI has gotten out of their efforts thus far.
Admittedly, the highest performance gains come from low resolutions without antialiasing. It stands to reason that the more CPU limited a test is, the more benefit the game will get from freeing up CPU resources. The real benefit to end users if only lower resolutions benefit is questionable, but every step helps. With the future of computer hardware firmly planted in parallelism, the burden of improving performance shifts a little further towards software developers. Coming up with new and interesting ways to parallelize code efficiently is going to be quite a new task for desktop software programmers to tackle. And in the end, Amdahl's Law reminds us that we are still limited by the benefit we can get from parallelism. The percentage of code that must remain sequential will become the limiting factor. But every little bit of parallelization still helps.
There are really quite a few questions to be asked about this driver. After adding up everything we wanted to do, the sheer number of tests we had laid out was enormous. In an effort to be more efficient ourselves, we decided to break our analysis of the 5.12 driver up. This article is meant as a quick look at the benefits of ATI's dual core enhancements on a few select games running on X1K series hardware. We will compare this driver to the old one as well as dual core performance to single core performance.
Our next look at the 5.12 driver will include a comparison to NVIDIA performance in single and dual core systems, more than one ATI card, more games, and as many more things as we can pack in. Of course we are open to suggestion. But for now, we'll take a look at what we've got to work with.
Facebook
Twitter
RSS
56 Comments
View All Comments
bldckstark - Monday, December 5, 2005 - link
The gentlemen at Sanda National Labs don't see the parallelization "problem" the same as most of the computing community, and take exception to Mr. Amdahl's observation. They have shown that the possible multiple CPU parallelization improvement can exceed Amdahl's theoretical maximum. They state that as the number of processors increase, the size of the problem introduced to the system increases allowing speeds above those previously thought possible. Here's a link to the horribly boring article.http://www.scl.ameslab.gov/Publications/Gus/Amdahl...">Extremely Boring Paper With Too Much Math
Thank God we got that cleared up. Now maybe I'll be able to sleep tonight. 8>}
phusg - Monday, December 5, 2005 - link
Interesting review, although disappointing performance.Just commenting to say don't forget about us SMPers! All the talk is about DC although not everyone with a dual CPU has it on the same die!!! I have a dual 760MPX board with a Radeon 9600 Pro and am hoping these driver improvement will really help me out.
Humble Magii - Monday, December 5, 2005 - link
This article is silly....Seriously I have been a long time Anandtech reader and this past year I have noticed a huge dropoff in quality! What ever happened to Anand's in depth reviews of CPU's and their architecture? I don't see that type of enthusiasm or detail anymore here.
Are the specific reviewers for certain articles biased? It seems this is the case whoever reviews a certain product nowadays is biased on this site. The new ATI drivers are crap not to mention are in beta.....
Who the hell cares about a driver update I mean seriously unless it was a huge change for the better or worse why report it? I don't see reviews on forceware changes or an AT article on the dual core drivers for Nvidia? I suppose there are no pro nvidia people at AT or at least maybe Nvidia doesn't pony up to AT's reviewer here?
Who has problems with Nvidia's newer drivers? I sure as hell haven't and I have a 4800+ with two 512mb 7800 GTX's. I never had problems with Nvidia's drivers. Everyone has a different system and just because you yourself have a problem when the majority do not does not mean the product sucks. I have had issues with a bios update utility due to beta creative drivers for a sb audigy I mean please people I would have never of thought of that but most problems happen because of what other software you have installed.
Who cares about the ATI X1K series I don't and most of the world doesn't especially since they are replacing it supposedly coming soon next year so why bother? They aren't even cost effective performance solutions comapred to Nvidia. Oh I would love to try a crossfire solution too bad there are zero around so guess what the only performance in graphics today is Nvidia. That's what we are stuck with now.
I swear this site no matter the trash ATI puts out will post it and put it in a better light. The forums here are even worse with all the newbs and their uneducated hate and bias to other solutions.
The site is dying Anand please replace Derek Wilson and some other members and start taking control of the reviews again everyone sorely misses you :(.
heulenwolf - Monday, December 5, 2005 - link
I have to agree on the point that since this driver update provides limited improvement, does it really warrant all the work Anandtech is putting into it. I own graphics cards from both ATI and NVIDIA and tend to go for midrange cards, so I can see some relevance to similar users who happen to own an x1000 series card. At the same time, as the previous poster points out, the advantage from a change of cards is far greater. So why so much testing that it has to be broken up into multiple reviews for a single driver providing such a small performance gain to users of a single line of cards who run at low resolutions and only play certain games? The benefit just seems small for so much work unless Anandtech can substantially tie it to a more general problem. My recommendation to the author, then, is to spend a little more time describing why you did all the testing - what it tells us - up front and tie it to a more general problem users may have.bob661 - Monday, December 5, 2005 - link
So the WE, the user, can know that this driver only provides "small performance gains". How else would you know?bob661 - Monday, December 5, 2005 - link
Damn it.....So THAT we...heulenwolf - Tuesday, December 6, 2005 - link
So, now we know. Where's the need for a follow up?hondaman - Monday, December 5, 2005 - link
I think this review is quite interesting. There is a lot of chatter in the channels about how DC is useless for gaming. Its reviews like this that give us guys with dual core cpus hope that our cpus are good for more than just playing mp3's and encoding movies all at the same time.DrZoidberg - Monday, December 5, 2005 - link
well after reading this article i still feel DC is only slightly useful when u are solely gaming. I mean why pay $200 extra for 2nd core to get 5% benefit at low res and when gaming at high res the benefit drops to like 2%. Like x2 3800+ and 3200+ are both 2ghz and about $200 difference and with these dual core drivers the 3800+ wins by 5%.Rather put the extra $200 to video card if main purpose of comp is to game and improvement will be like 30%+, like going from 6600gt to 7800gt. However multitasking is a totally different story. Encoding movie and gaming at same time, then dual core is very worth it.
porkster - Monday, December 5, 2005 - link
You may only get 5% as the games are not taking advantage of the new features.