Sign in with
Sign up | Sign in
Your question

Same Graphics card, different PC, different results?

Last response: in Graphics & Displays
Share
October 30, 2006 9:39:15 PM

I have a question regarding the requirements of a system for frame rate. I recently got Half Life 2 and checked the frame rate between two machines.

machine a)
HP
Graphic Card - ATI 1900XT PCI-Express
AMD 64x2 Dual Core 4400 - 2.2GHz
1GB RAM
Window XP Media Edition
Frame Rate: 74fps HL2 Episode One test

machine b)
Dell
Graphic Card - ATI 1900XT PCI-Express (same exact physical card)
AMD 64x2 Dual Core 4800 - 2.4GHz
1GB RAM
Window XP Home Edition
Frame Rate: 125fps HL2 Episode One test

Now the only difference I see is the different AMD core (2.2GHz vs. 2.4GHz), but does it really make that much of a difference between the 2 systems or am I missing something else? I didn't do any tuning for the card via either the game itself or the card, maybe there is something in the game I can move, like lower some resolution or something, that my friend did to machine b (and isn't telling me), which I didn't do to machine a. Is there any ideas someone could share that may make sense, other than buying a new CPU, and finding out it gets the same results. If power load was the issue I'd assume it would just no even work.

- Waffles.... WHUT -

More about : graphics card results

October 30, 2006 10:08:35 PM

The slower system is most likely suffering from a software issue.
October 30, 2006 10:14:39 PM

Before running each test, ensure the graphical settings are equal (for ease just set them to Default). Also, are both machines using the same driver?
The RAM might be of different spec also, that's something you need to check; but I don't see RAM making such a large difference.
The most obvious thing that I can see being different in both these setups would be the resolution the test is running at. Try it again, and ensure that the graphical settings, including resolution, are set to the same values, and run the test 3 times, then average to ensure a more accurate result.
Related resources
October 30, 2006 10:29:05 PM

You can put most of your money on the processor really. I have a 1950 in on a conroe 6600 and im doing 150 fps in the Lost Coast test. I'm also sporting 2 gigs of ram. Don't be fooled, anything that can speed the system up can easily change the performance.
October 30, 2006 10:35:48 PM

200 mhz cpu speed is not going to cause an 80% framerate increase. My money is on driver, background load, game settings.
October 30, 2006 10:36:48 PM

true, but it would be nice to dream :D 
October 30, 2006 10:52:39 PM

XP Media Center is XP Pro with media center software.

It will use the same resources as XP Pro, which if configured properly is no more than XP Home.

This of course is assuming you don't have media center running while gaming :) 
October 31, 2006 1:08:49 AM

Are you sure one pc is not overclocked?
a b U Graphics card
October 31, 2006 2:28:11 AM

Are you running the EXACT same test, such as a pre-recorded ingame test like the one in FEAR. If you simply play the game on one and then play it differently on the other then you will almost definitely see a difference in performance - especially if one test has many more effects than the other.
October 31, 2006 2:59:59 AM

Apples and oranges man thats what this comparasin is. :o 

The systems most likely have a totaly different configuration. along with alot of other things like were stated already that can be causeing that. Its highly unlikely that the cpu is changing things more then 5% do you even know the resolutions video settings running software on both machines? IMO there is no way your going to get a equal comparasin from those two machines especialy since they are OEM that hardware ehh forget it. There is alot more different about those systems then just the video card. 74frames though isnt bad its more then your eyes can see.
November 9, 2006 3:49:43 PM

I re-checked the graphics settings. The resolution was different, but I found it didn't make a huge difference. I originally thought I was running the same test, but turns out I wasn't:
HalfLife2 Lost Coast Video Test (option in menu before games are started)
vs.
HalfLife CounterStrike Video Test (option in menu before games are started)
Give very different frame rate results.

After <psuedo> apples to apples test the difference was less.

2.2GHz
ATI 1900
HL2 Lost Coast Menu Video test
Frame Rate: 83fps
Video
resolution 1024x768
Full screen
4:3 aspect ratio

vs

2.4GHz
ATI 1900
HL2 Lost Coast Menu Video test
Frame Rate: 98fps
Video
resolution 1024x768
Full screen
4:3 aspect ratio
------------------------------
Resulotion didn't make too much of a difference, 2-3 frames. Someone mentioned that you couldn't tell since it is more than 30 frames. In theory that should be absolutely true, but I do notice that the 98 is more fluid and the 83 seems to jump frame. This may not be a result of the frame rate, it could be Steam or something burning up CPU, but this frame jump every 5-15 seconds was noticable on the slower frame rate. I appreciate all the good input, next time I'll do more diligence testing up front, not test the wrong thing.

I have the F.E.A.R. game loaded on the slow machine. I'll try it with the fast machine/good card, but are the graphic that good? They kinda suck compared to HL2, even on my slow PC. I guess thats a discussion for a different forum.

- WAFFLES OUT -
November 9, 2006 5:51:20 PM

Quote:
You can put most of your money on the processor really... Don't be fooled, anything that can speed the system up can easily change the performance.


This is totally, absolutely, and completely wrong.

The difference between a 4400+ and 4800+ CPU is almost totally meaningless. If there is a 5% difference in framerates because of the CPU I'd be surprised.

Newer games like this are bottlenecked by the videocard mostly. Even in older games though, the difference between the 4400+ and 4800+ will be minimal.

At the highest playable resolutions and settings, where the card is bottlenecking, even a Conroe won't make a noticable difference over an Athlon X2.
a c 358 U Graphics card
November 9, 2006 11:56:42 PM

Some of the performance difference could be due to different chipsets being used. But typically that should only account for about a 5% difference.

Perhaps your systems are not fully optimized. It's surprising how preformance can increase with a fresh Windows XP install.
November 10, 2006 3:04:35 AM

Tecnicly people can see the difference in frames up to 60. For me when it starts to go under 60 i start to see the frame problem. at 30 its just so much snap shot crap i cant play. Course resolution does make a big difference in frames having to render all those extra pixles takes some power.
November 10, 2006 2:38:42 PM

Do you think going from 1Gig RAM to 2Gig or 3Gig would help the frame rate any? Or is the bottleneck in the GPU and video card memory area? What function does the video card use motherboard RAM for? Is there a good place to look at video card pipleline/video path information?
November 10, 2006 3:08:43 PM

Over 1 gig of RAM makes almost no difference in framerates in 95% of titles, although it does speed up load times between levels alot.

Some titles, like the new Battlefield games, like alot of RAM and will perform a bit beter with 2GB. But they are the exception, not the rule.
a b U Graphics card
November 10, 2006 11:10:58 PM

I agree, but games like crysis (when it comes out) need a minimum of 1gb with vista, so 1.5-2gb could soon become the new standard. Going more than 2gb is pointless tho.
!