[HD5870] Unstable FPS in older games

[tonysathre]

I still enjoy playing the old Call of Duty: United Offensive game online from time to time. I usually cap my frame rate at 333, but it seems to drop down a lot, to around 140-180 depending on the amount of players, and the size of the map. On some maps, it even drops as low as 125. I've talked to other people that have much older hardware than me, and they seem to be able to hold a constant 250, which makes me think that I should be able to handle a little more than that considering my hardware is much better. I would have assumed this beast of a card would have been able to handle such frame rates steadily on such an old game, but maybe I'm just wishful thinking. Can anyone shed some light on whether it should be able to handle such frame rates, or are those numbers just unrealistic? Below are my specs:

Intel Core i5 750 @ 2.8 Ghz
ATI HD 5870
4 GB Corsair XMS3 1333 CL7

If you need anymore info just let me know.

Thanks,

Tony

 

[Maziar]

How is the situation in other games ?

 

[tonysathre]

The only other game I've tried is Battle Field: Bad Company 2, and I could run that at max settings just fine, but I never checked the FPS. The FPS in those games doesn't matter as much as in Quake 3 engine based games.

 

[jerreece]

As long as modern games like BC2, or Metro 2033 would run with good FPS and performance, I would absolutely not worry about older games.

Especially if you're seeing 125 to 250 (or more). It's all pointless that high anyhow. The only reason you're noticing anything is because you're checking the FPS meter.

Keep in mind however, that internet based games are affected by Latency (even FPS can be affected by latency issues). This is often times where folks with games like WoW complain about poor FPS in cities even though they have powerful machines. If your computer is sitting around waiting for data transmission it may pause slightly as it renders new scenery, thus lowering FPS over all.

Where as in a more modern game like BC2, there's plenty of stuff to render that it keeps busy even with poor latency.

 

[tonysathre]

Especially if you're seeing 125 to 250 (or more). It's all pointless that high anyhow

Um, yes it does matter, in Q3-engine based games. It makes a huge difference. Read this:

http://www.funender.com/quake/articles/strafing_theory.html

I would absolutely not worry about older games.

I play more older games than newer games. Too many new games don't rely on pure skill anymore (aside from QuakeLive). They have unlockables, perks, and other things of that nature, that make them too easy.

I just tested it with Left4Dead. With everything on max settings I get 190-250 constant FPS. Is it that older games can't utilize all the potential power of newer hardware?

 

[bystander]

Especially if you're seeing 125 to 250 (or more). It's all pointless that high anyhow

Um, yes it does matter, in Q3-engine based games. It makes a huge difference. Read this:

http://www.funender.com/quake/articles/strafing_theory.html

What does this artical have to do with FPS?

And to really disapoint you, the maximum FPS you actually have, is likely 60 FPS, because that is the most the vast majority of monitors will display today.

 

[amk09]

What does this artical have to do with FPS?

And to really disapoint you, the maximum FPS you actually have, is likely 60 FPS, because that is the most the vast majority of monitors will display today.

True statement, that article is a "Theory" btw.

Being unhappy about 125 fps is absolutely absurd.

 

[tonysathre]

If it didn't matter, I wouldn't be asking. Obviously you have no idea how the Quake 3 engine works.

If you can't answer my questions, or don't have a solution to my problem just say so. Simple as that.

What does this artical have to do with FPS?

Maybe this was more suitable to the situation:

http://www.funender.com/quake/articles/fps.html

Tony

 

[bystander]

Your first artical was about strafe running, which many engines have been adjusted so it is not an advantage.

The 2nd one, is just some calculation that doesn't explain much and as I plug in different framerates, it rarely changed results much.

Having high FPS did not give an advantage, having a FPS line up on a certain multiple did.

 

[bystander]

You also might try a higher clock on your i5. With after market air cooling, an i5 should easily be able to get up to 3.3-3.6ghz

 

[bystander]

Btw, what monitor resolution are you using? If you really insist on trying to gain some magical FPS, a lower resolution will likely fix your issues.

 

[jyjjy]

This isn't a problem. A frame rate higher than your refresh rate is entirely useless. For the smoothest experience(and to prevent screen tearing) you should just be turning on vsync for games that are consistently higher than your refresh.

 

[tonysathre]

Having high FPS did not give an advantage, having a FPS line up on a certain multiple did.

That's my whole point. 333 is the best of the "magical" frame rates.

which many engines have been adjusted so it is not an advantage.

Yes, but this game still runs on it, as does all the Call of Duty games up until CoD:MW2.

I've tried playing at 800x600, 1024x768, and 1280x1024, and they were all pretty much the same.

What I can't figure out, is that this game is almost 7 years old, and it runs worse than newer games. Why am I getting better frame rates in newer games, compared to the older ones? I'm running this old game at medium settings, and newer games at max settings, and getting the same frame rates. It doesn't make sense.

You also might try a higher clock on your i5. With after market air cooling, an i5 should easily be able to get up to 3.3-3.6ghz

I've already planned on doing that soon. My motherboard autoOC'd it to 3.7 and it ran fine, but got hot under high load.

This isn't a problem. A frame rate higher than your refresh rate is entirely useless. For the smoothest experience(and to prevent screen tearing) you should just be turning on vsync for games that are consistently higher than your refresh.

...

 

[bystander]

You also have to realize that the games you are running are designed for older hardware. Newer hardware may actually hinder performance.

And you might also consider that the difference you'd gain for hitting 333, is so minutely small, if it exists at all, that you'd never notice a difference anyways.

 

[amk09]

This isn't a very logical explaination but maybe it's because the 5870 isn't designed for games that are 7 years old. If you have an older card try swapping them out and see if it does anything. I'm more than likely wrong though.


EDIT: sorry bystander I didn't see that you posted that already, nice hivemind

 

[yannifb]

Those old games, like jenny said, are probably not able to properly utilize the 5870. Also they are probably only using one thread/core on your i5. Games nowadays use 2-4 which actually makes a difference, especially in strategy games.

 

[tonysathre]

Ah, I see now. Thanks for all the responses. That does make sense.

 

[tonysathre]

Best answer selected by tonysathre.

 

[amk09]

Ah, I see now. Thanks for all the responses. That does make sense.

Tony, do you have hyperthreading enabled?

 

[jyjjy]

This isn't a problem. A frame rate higher than your refresh rate is entirely useless. For the smoothest experience(and to prevent screen tearing) you should just be turning on vsync for games that are consistently higher than your refresh.

...

The refresh rate is how many frames your monitor can actually display per second. Vsync is used to sync the frame rate of the game to what your monitor is actually displaying. Frame rates higher than your monitor can display are necessarily totally pointless.

 

[tonysathre]

Tony, do you have hyperthreading enabled?

I'm not sure. Would that be in the BIOS?

The refresh rate is how many frames your monitor can actually display per second. Vsync is used to sync the frame rate of the game to what your monitor is actually displaying. Frame rates higher than your monitor can display are necessarily totally pointless.

I know what vsync and the refresh rate do. If I turn vsync on, my frame rate gets capped at 60, which is 60hz, my monitors refresh rate. Going by what you're saying, explain why the game physics/mechanics change when you set your max frame rate at a certain level.

 

[amk09]

Ah, excuse me I meant Turbo Boost, not hyperthreading, as the i5 750 doesn't support hyperthreading.

But as for Turbo Boost, you would access this through the BIOS settings, and what it does is allow you to control the number of active cores on your CPU, so essentially you could make your CPU into a single core processor(or dual/triple core), which may in fact help you out with the dated game your playing.

Try enabling Turbo Boost and let me know how that helps.


EDIT: I would like to apologize furthermore for my lack of knowledge, it seems that you don't need to go into your BIOS to enable turboboost. All you need to do is go to processes in windows task manager and choose Q3 and set its affinity to however many cores you want.

 

[jyjjy]

explain why the game physics/mechanics change when you set your max frame rate at a certain level.

I highly doubt they actually do so my guess is placebo effect. If they actually do then that is quite poor coding, especially for an online game where the physics/mechanics of game play need to be identical for all the players.

 

[tonysathre]

EDIT: I would like to apologize furthermore for my lack of knowledge, it seems that you don't need to go into your BIOS to enable turboboost. All you need to do is go to processes in windows task manager and choose Q3 and set its affinity to however many cores you want.

Ah, yeah I've tried that before I posted but it didn't seem to help. Thanks though.

I highly doubt they actually do so my guess is placebo effect. If they actually do then that is quite poor coding, especially for an online game where the physics/mechanics of game play need to be identical for all the players.

And it's that "poor coding" as you call it, that made iD Software, and the Quake series a legend.

I think I've got my answer. Thanks to everyone that constructively replied.

Tony

 

[jyjjy]

My answer was placebo effect, not poor coding actually but yes, a physics engine for a online multiplayer game that relies on the particular FPS of a certain client is most certainly poorly coded.