Sign in with
Sign up | Sign in
Your question

Performance new pc somewhat dissapointing..

Tags:
  • Homebuilt
  • Performance
  • Systems
Last response: in Systems
Share
September 6, 2006 7:15:18 AM

Hi there!

I hope someone can help me with a bit of information.

Last week i build my new pc hoping to play all my games at max settings. But I´m somewhat dissapointed..
I cannot play Quake 4 multiplayer at a steady 60 fps.. I have everything at max settings though. 1280x1024 ultra quality, Ati driver settings at optimal quality.
when there are quite a few players my fps can go down to 30-40..

same thing for F.E.A.R. combat.

Is this normal for my rig? Am I really already pushing it to the max?

specs:

Intel E6600
Asus P5W DH Deluxe
Corsair TWIN2X2048-5400C4 2048 MB
Sapphire Radeon 1900 XT 512 MB
Maxtor DiamondMax 10 - 300 GB

At first I thought I had some settings wrong, but everything is set correctly. Also ran some benchmarks:

3dmark01 - around 40.000 points
3dmark03 - around 20.000 points
3dmark06 - around 5.700 points

Seems quite ok right?

Was I too hopefull for this rig?
Thx in advance!

More about : performance dissapointing

a c 90 B Homebuilt system
September 6, 2006 7:22:26 AM

Run the FEAR built in benchmark test and report those results. Those are the ones you need to measure against other systems benchmarks. The multiplayer results can be too easily skewed by poor connections, server or high latency issues.

This Tweak Guide explains how to benchmark Quake 4.
http://www.tweakguides.com/Quake4_10.html
September 6, 2006 7:34:18 AM

Thx WR2! I´ll run them after work and report the results.
Related resources
September 6, 2006 10:25:25 AM

the problem may lie with your internet connection.
Your connection may be slow.
When playing FEAR Combat note the ping times.
If you get very high ping times that is what is affecting your performance.
the fact that when there are more players the performance decreases substantiates this claim.
September 6, 2006 11:17:30 AM

Hi Blade,

I don´t believe my connection is the problem. I´ve got a 4000/600 kbit/s cable connection and my ping is around 30-40 so that should be fine.

My fps go down to 30 fps on a busy q4 server. I was hoping someone with a similar rig could tell me if they also have these framedrops.
I really thought i bought a nice gaming pc to play quake 4 multiplayer at ultra setting at 1280x1024 resolution..
September 6, 2006 11:29:37 AM

Try quake 4's single player and see if you get the same framerate issues. Because quake 4's net code is just as bad as Doom 3's. And your framerate beleive it or not is highly dependant on the server it self. Im downloading FEAR combat ill be able to tell you if I have the same issues.
September 6, 2006 11:45:37 AM

Thx Mike, I didn´t know q4´s netcode was so bad. I´ll try quake 4 single player at max settings as soon as I get home from work. Hopefully it´ll run at a constant 60 fps! :) 
September 6, 2006 11:54:58 AM

good luck with FEAR combat my friend ussualy can play FEAR SP with effects on max and threst on high/medium... with good fps.. but when it comes to FEAR combat the ping in most servers is so high and also the fps isnt the best :( 
i am going to download it when i get back home... its so sweet :) 
September 6, 2006 12:11:53 PM

Since when was 30-40fps low...
If your framerate never drops below 30 at maximum settings, I think you've done rather well for yourself, kwan.
September 6, 2006 12:45:10 PM

well after playing quake 3 at 125 fps for years, 30 tends to feel a bit slow..
I think I just have to get used to it, or trade in some eye candy for a steady 60 fps. Just didn´t think that was neccecary with my new pc.
September 6, 2006 1:26:56 PM

i just played the demo on max and it ran just fine :) 
goodluck.. i have also heard that Quake 4 ussualy runs better on Nvidia cards.. but i dont know if its really true
September 6, 2006 2:25:39 PM

Is this 30fps you're talking about the minimum you're dropping to? If so thats probably pretty good for multiplayer where the game must track other players random behavior etc rather than the single-player game. Most games are benchmarked using a timedemo or other such taken in single-player like environment which would alone improve frames reported usually.

Possibly the most important thing is, Does this dip to 30 frames feel sluggish or lag you? If it never falls below 30fps then lag is likely your connection as others have mentioned, latency etc. If it does indeed fall below 30fps at times it could be combination of the two.

Also some benchmarks you see on sites report max frames when they're in the 100s and/or average frames (70s or so usually depending on card with max settings if I recall correctly). Also if you have v-sync enabled it will limit your top end fps although this won't affect your bottom fps it can give more "headroom" for top end if you like benchmarking your stuff. Just something to think about when comparing your own bench to some on the web as its almost never apples to apples :) 
September 6, 2006 4:09:46 PM

Quote:
Most games are benchmarked using a timedemo or other such taken in single-player like environment which would alone improve frames reported usually.


doh didn't realise that.. silly me.. guess I was indeed comparing apples to bananas :D 

30 fps is indeed the minimum i'm dropping to in multiplayer max settings. single player is almost constant at 60 fps so my system is probably working as it should be.

Thx for the help guys!
September 6, 2006 6:21:12 PM

Yup your system is working fine, I think you are expecting too much from a single GPU setup.

If you are running at 1280x1024 and above you would normally be fine, but when you start adding 4xAA and 8 or 16x AF then this is where SLI and Crossfire begin to show their strengths.

If you want to scrape a little extra grunt out of the X1900XT, then I'd suggest sitting an Accelero X2 on it with Arctic Silver 5, and overclocking to XTX speeds (entirely at your own risk of course). The Zalman VF900cu is a better choice for a quiet setup at low speed, but at the highest fan speeds (equivalent cooling to the Accelero for overclocking) it is actually louder.

Even though it is a very fast CPU anyway you would also get an improvement from overclocking the E6600. To not even have it at E6700 speeds is just plain wrong. That has to be one of the easiest ways to save $150 ever.

Oh, and a quick defrag of your hard disk can work wonders in keeping a consistent framerate... particularly because your Maxtor isnt the fastest around!
September 6, 2006 6:34:21 PM

Hey Dave, thx for the advice. I'm not so keen on overclocking so I've decided to start saving for another X1900XT and use Crossfire :D  but I have a OCZ 450 Watt power supply, that probably won't cut it will it?

About the hard disk, which one is a good one for gaming? Does a raptor 10.000 rpm for examply make that much of a difference?
September 6, 2006 6:47:58 PM

I just played a bit of FEAR Combat and my lowest Framerate was 40, my highest was 96. Thats with all settings set to maximum, 1280x960, with 4xAA, and 16x AF. Also I have a question are you using temporal antialiasing ? because that destroys framerate. A raptor wont make much of a difference, I can load games faster then 2 of my friends who have raptors, and I just have a regular IDE Western digital 80 gb. Your powersupply might not be cutting it atm, 450 watts using a x1900 xt is a little close.
September 6, 2006 6:52:03 PM

A Raptor will make more difference to load times than anything - don't expect a great hike in FPS if you have 2GB of RAM.

However, I'm certain I'll be including one in my next build as people's opinions of them tend to be very good; the enhanced loading times for games and your OS seem to be worth the added investment.

And yes, that OCZ should handle one X1900XT (ATI recommend a 450w PSU minimum for this card) but you will probably need to upgrade it for a crossfire setup. Not just because of the wattage but because of the current requirements on the +12v rail(s).
September 6, 2006 7:02:51 PM

hmm so that's gonna be a new gfx card, power supply and hard disk.. lol I wish I stayed away from pc's.. it's costing me a fortune!

Mike, no I use regular antialiasing. but seeying that ur fps also drop down to 40 confirms it once more that my rig is working fine. I just need Crossfire :D 
September 6, 2006 7:39:12 PM

I have the EXACT same rig, and was expecting a little more performance when playing FEAR Combat. I see some guys on there destroying other players, and I'm convinced they have faster rigs (although I havent played a FPS in about 5 years). AHHAHAHAHAHAH.

How do I really rack up my performance in that game? Where is this built in benchmark program someone ealier was talking about? Please let me know what my settings must be set to for MAX MAX MAX performance. THANKS GUYS!

FYI: Here are my specs
http://forumz.tomshardware.com/hardware/modules.php?nam...
September 6, 2006 7:48:48 PM

Your results sound pretty nice to me (look at sig for somewhat outdated rig). I'm actually going to build an almost identical system to yours in about 1 month and I'd love to "dip" down to only 30fps hehe. Enjoy your rig :) 
September 7, 2006 12:38:16 AM

Quote:
I have a OCZ 450 Watt power supply, that probably won't cut it will it?


Won't. Don't think about it. I also believe you would be better of with 7950GX2 than with CF or SLI setup.

Quote:
Does a raptor 10.000 rpm for examply make that much of a difference?


As much difference as between the numbers 7,200 and 10,000 percent-wise. Another good one is WD4000YR with its sustained speed of 65MB/sec.

Perfect Disk defragmenter can really help in case your HDD is heavily fragmented.
September 7, 2006 2:02:22 AM

Don't worry about getting a raptor. It will only effect loading time and that too by not much.

If u can trade ur 1900 xt + some cash for 7950, then do it. Otherwise save your money for another 1900 xt and a better psu.

U could also wait for dx10, buy 7950 equivalent and sell ur 1900xt.
September 7, 2006 4:53:45 AM

So which is better? A single 7950 or X1900XT crossfire? also, if i sell my X1900XT and buy a single 7950, will my 450 Watt supply handle it?
September 7, 2006 1:46:47 PM

Hi Guys,

I ran PCMark and 3DMark on my PC last night. I was using the free version. As I am not using the pay service (just not worth it for one run at $40 if you dont use it professionally), I cannot check to see how my system stacks up against others. I hit Google and the forumz, and got some data points. If those points prove true, then I think I am going to be very disappointed with the new system. Could you please take a look at my scores and let me know what you think? Please post yours too so others coming to this page can learn too. THANKS!

PCMark = 6493
3DMark = 4936

The list of detailed components are on this page:
http://forumz.tomshardware.com/hardware/modules.php?nam...

-Antec Performance One P180B Mid-Tower
-Antec NeoPower NeoHE 550 550W
-Asus P5W DH Deluxe
-Conroe E6600
-OCZ Platinum Revision 2, 2GB, DDR2 800
-MSI RX1900XT-VT2D512E Radeon X1900XT 512MB GDDR3
-Seagate Barracuda 7200.10 ST3320620AS 320GB
-Lite-On SHM-165H6S Dual Layer DVD±RW Writer
-ViewSonic VX922 19" LCD Monitor
-Microsoft Wireless Optical Desktop Elite Mac/Win PS2/USB
!