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
Pjotr - Tuesday, December 6, 2005 - link
Why not simply show a screen shot of Task Manager after each benchmark? Then we can see approximatly how much of the second core is used by each benchmark.SemiconductorSlave - Tuesday, December 6, 2005 - link
If you follow the link Amdahl's law in this artilce and then at the bottom of page follow link “Reevaluating Amdahl’s law” you see the author state,Amdahl's law contains ” . . . the assumption that p is independent of N, which is virtually never the case. One does not take a fixed-size problem and run it on various numbers of processors except when doing academic research; in practice, the problem size scales with the number of processors. When given a more powerful processor, the problem generally expands to make use of the increased facilities.”
Isn’t this also what we have seen with video games, that they have always expanded to make use of the increased facilities?
SemiconductorSlave - Tuesday, December 6, 2005 - link
Derek's article mentions in its conclusion, “The real benefit will come in when game developers start working on parallelizing their code as much as possible.” This article is very forgiving of ATI’s new driver as there is already significant benefits to dual core processors on Quake4 and SeriousSam2 if you are using Nvidia 81.xx drivers.The article, using the ATI X1800 XL with the Catalyst drivers, Quake4 showed “no performance difference or issue” in the single core to dual core tests. In an article on www.xbitlabs.com titled Contemporary CPUs and New Games: No Way to Delusions!, on page 6, Quake 4 is shown to run much faster on dual core using the Nvidia 81.xx drivers, as the X2 3800+ clocked 2 GHz gets 101.6 fps, and the 3500+ clocked higher at 2.2GHz achieves only 98.6 fps. The 3200+, which is clocked evenly with X2 3800+ at 2Ghz, only achieved 93.4 fps, which means the dual core produced an 8% gain.
Also, on page 5, Serious Sam2 is shown running faster on dual cores, as the X2 3800+ clocked at 2 GHz achives103.1 fps beating out the 3800+ clocked at 2.4 GHz which achieved 99.2! The 3200+ clocked evenly with X2 3800+ at 2Ghz only achieved 84.8 fps. This indicates the dual core produced a 17.7% gain! The author notes on this test, “I have to point out NVIDIA drivers also started supporting dual-core architectures. ForceWare version 81.xx allows enjoying the advantages of dual-core technology in DirectX as well as in OpenGL.”
And to rule out that having two times the cache on an X2 is the reason for the performance difference, you can directly compare the equally clocked 4000+ to the X24600+. Each are clocked at 2.4 GHz and have a total of 1Mb cache. In Quake4 the X2 gets 110.2 while the 4000+ only achieved 103.2, which is still a 6.35% difference.
I think it would be great if this article is appended or these facts are included in the update. The end user should know what dual core performance is available to them now, not think that because ATI wrote a less than successful driver that we have to wait for the game developers before we see any significant dual core benefits!
Thanks,
Semiconductor Manufacturer and Anandtech fan.
porkster - Monday, December 5, 2005 - link
Ok just read the article. I thought the test was to be scheduled. Anyway, why are you only testing on an AMD system and why no sign of Black&White2, one of the new era of games that will push the bus, system and visual experience? Also Intel are the kings on multitasking and bus bandwidth so why test dual core drivers and speed % differences on an AMD?Surely a driver that has two cpu threads going will require more bandwidth on the bus when expected to move upto twice the amount of graphic files around tot he gfx card.
Can you please test a game that is modern like B&W2 in the future and put at high res. Compare the benefits between the market products. Show results on bandwidth demand and their effect whilst multitasking the system.
It seems Anandtech is falling behind the times due to bias towards AMD. It maybe ok to fool the majority of system owners that still have low PC-memory ratings and legacy AMD stuff thinking they're modern, but you portray your site has a leading tech reviewer. Please show the best of, not best of frame rates for non multitasking environments running an old tech wise game.
DrZoidberg - Tuesday, December 6, 2005 - link
Most of the readers here that have dual core systems have AMD X2 processors. AMD is faster at games than Intel so it is totally appropriate for AT to test using AMD first. AMD is not legacy, its more the general public still view Intel as the performance leader based on their processors several years ago and still clueless that AMD have now taken over.
Sure AT should now test using Intel processors just for knowledge sake but i bet performance % is similar to AMD like 1%-10%.
Quake 4 is new just came out recently, DOD source is recent, Battlefield 2 is a popular game many people play it, B&W2 would be nice but COD2 is the game really missing here.
porkster - Tuesday, December 6, 2005 - link
Actually you maybe wrong regarding AMD best for gaming, as an example compare the results on this test to http://www.anandtech.com/mb/showdoc.aspx?i=2631&am...">http://www.anandtech.com/mb/showdoc.aspx?i=2631&am...DrZoidberg - Tuesday, December 6, 2005 - link
Why have u provided a link showing motherboards for Intel at games and no AMD motherboards?What does that prove? That one particular Intel chipset motherboard is better than another Intel motherboard?
This is a more appropriate article: http://www.anandtech.com/cpuchipsets/showdoc.aspx?...">benchmarks
Battlefield 2
Pentium D 830 (3ghz) 92.1 fps
Intel Yonah (upcoming Intel processor) 103.3 fps
AMD X2 4200+ 120.8 fps
porkster - Wednesday, December 7, 2005 - link
Compare the results for the Farcry from this test results to the one I linked to.The Intel system is far better yet it's not using the dual core style gfx drivers yet.
I'm starting to think Anandtech doesn't want to compare latest games between the best of AMD and Intel as Intel will show it's better, especially at maintaining FPS during multitasking.
SemiconductorSlave - Wednesday, December 7, 2005 - link
Like in these 4 benchmarks?http://www.anandtech.com/cpuchipsets/showdoc.aspx?...">http://www.anandtech.com/cpuchipsets/showdoc.aspx?...
"We have Firefox loaded with all 13 tabs from our new suite test, iTunes is running and playing a playlist, and Newsleecher is downloading headers. We kept Newsleecher in this test simply because it's the best way for us to be able to have a fairly CPU/disk intensive downloading task running in the background while still maintaining some semblance of repeatability." --Anand Lal Shimpi
porkster - Thursday, December 8, 2005 - link
TALK ABOUT OLD REVIEW. Sorry, be relative.