Very Low FPS with Radeon 9700 Pro

[SLCentral]

Hey,

I am benchmarking for a review on the 9700 Pro, but I'm receiving unusally low benchmark results in:
Comanche 4 (41 FPS)
Quake 3 (240)
RTCW (111)

All are tested on 1024x768 resolution on "High Quality" settings, and vSync is OFF.

My specs are:
Athlon XP 2200+
Asus K7N266-C
512MB DDR 2100
ATI Radeon 9700 Pro

I'm using Catalyst 2.5, but I've also tried 3.0.
And yes, AA and AF is OFF.

Please help me out, I need the review ASAP.

Thanks!

 

[onza]

just out of curiosity, did u have any other brand name video card in the system b4 your ati card?

 

[SLCentral]

I had a Ti 4600, but I formatted the computer after. And, I perform well in 3DMark03 (4400), and I get 12708 in 2001SE (don't know how good that is), as well as getting pretty good scores in UT2003.

 

[oLLie]

that's a little odd, i want to know what's going on too.

bump

 

[BFG10K]

Make sure you have the latest BIOS and chipset drivers installed for your system and disable all background programs.

 

[SLCentral]

There are no background programs running, and I am running the latest nForce drivers, as well as the latest BIO's for my motherboard. As I said before, other benchmarks such as 3DMark2001/2003 don't have this problem, so it has to do with the Quake engine, as well as the Serious graphic engine. I've got no idea what engine Comanche 4 is. If anyone has ANY solutions, PLEASE let me know!

Thanks,
Daniel

 

[Socio]

Damn it!

I guess I joined the club, my 4600ti died and replaced it with the 9700Pro and the 9700Pro is slightly slower in benchmarks for me to.
In fact I only got about 1 point higher in flyby and 2 points lower in bot with UT2003 benchmark compared to the 4600. I think we are victims of the ATI hype, no way is it as fast a the FX it's barely as fast as a 4600.

 

[Viper96720]

It's cpu limited and you do have vsync off. Benchmark it with a barton or 3.06 P4. At your cpu speed you won't notice much difference between it and a ti-4600. Until you enable aa and af then the scores don't go down as much a 4600 would.

 

[SLCentral]

Yeah, I wish I could use a Barton/3.06 , but, it's lucky that my review is looking mainly at anti-aliasing and AF peformence then regular performence.

So basically my performence is normal? Well that SUCKS! I just got my 9700 Pro and I was expecting awesome results, but, I guess not. However, playing Unreal 2 at 1280x1024 with AA and AF on is pretty sweet.

Oh yeah, VSync is off.

 

[Socio]

Originally posted by: Viper96720
It's cpu limited and you do have vsync off. Benchmark it with a barton or 3.06 P4. At your cpu speed you won't notice much difference between it and a ti-4600. Until you enable aa and af then the scores don't go down as much a 4600 would.

So what your saying is with AA and FSAA turned off the 9700Pro is not quite as fast as a 4600ti and with AA and FSAA turned on its going to be even slower just not as slow as the 4600ti would be with them on?

If thats the case why would anyone ever want AA and FSAA turned on? The more I read the more buyers remorse I get. Not only that but it would mean 9700Pro is not faster than either the 4600ti or the FX, it just does not slow down as much with everything on.

What a load of ATI crap!

 

[UncleWai]

Research before you buy, not the other way around, then you won't get buyers remorse.

 

[policy11]

What's the point of getting more than 125 fps in quake3?

http://www.quake3world.com/ubb/Archives/Archive-000001/HTML/20000822-3-015173.html

 

[Viper96720]

I meant it's faster than a 4600 with anti aliasing and anistropic filtering or without in certain benchmarks. Some benchmarks are cpu limited so depending on the resolution you won't notice much of a difference. Need to crank up the resolution to really see a difference.

not enabled
enabled

 

[Viper96720]

Your comanche scores seem right. Here are anandtech scores 49.7 at 1024x768, 49.6 at 1280x1024, and 47.2 at 1600x1200. This is from a 2.53ghz P4 PC1066rdram system

 

[ScrewFace]

I think your ATHLON XP 2200+ is the bottleneck on your system. If you had something like a Pentium4 3.06GHz overclocked to around 4GHz then you'd see your RADEON 9700 Pro shine.

 

[Socio]

Originally posted by: Viper96720
I meant it's faster than a 4600 with anti aliasing and anistropic filtering or without in certain benchmarks. Some benchmarks are cpu limited so depending on the resolution you won't notice much of a difference. Need to crank up the resolution to really see a difference.

not enabled
enabled

Those links prove the 9700Pro IS NOT FASTER, it just does not slow down as much with everything on.

The fact is with everything cranked to high performance, in all benchmark tests I have run so far the 9700Pro is slighly slower than my old 4600ti.

 

[SLCentral]

I agree with everyone that my CPU is holding it back, but why on FiringSquad (http://firingsquad.gamers.com/hardware/athlon9700/page4.asp), on a very similar system configuration, they get 248fps in Quake 3, while I'm getting around 210 now. It doesn't make a lot of sense. Sure, it may not be very noticeable, but it will probably stand out in my review, which may make it look like a bad card.

I'm going ahead with the review and just getting the data I got, whether it's normal or not. I'll just explain in my review that my CPU is holding it back, but, like I said, it doesn't make sense that FiringSquad is actually performing 30+ points better on the same game, while our system configs are the same. Would it have to do with the fact that I'm using the Demo of Quake 3 Arena?

 

[gururu]

the 9700 won't be much faster until the number of operations per frame increases; that is at low resolution with old games, the 4600 and 9700 might be neck to neck. Crank up the resolution alongside aa and af and there is no doubt the ati crushes the 4600. a lot of this is also cpu dependent. though you're cpu should be speedy enough, a faster cpu might improve the difference. But your cpu is good because it is more accurate with what most people have.

go with your results, but add some truly killer benchmarking such as ut2k3 at 1600x1200 with aa and af. use newer games, as Q3A is outdated. Comanche is a hog and might be more cpu dependent than the others.

good luck!

 

[kylebisme]

Originally posted by: Socio
The fact is with everything cranked to high performance, in all benchmark tests I have run so far the 9700Pro is slighly slower than my old 4600ti.

no, the fact is that your 9700pro is considerably faster than a 4600ti, but your cpu is still the same speed. what are you benchmarking for anyway? you have the best videocard on the market, go play games already and don't crank anything to high performance. if you do you don't desrve the card and you might as well just trade it off for a 9000 or a mx.

 

[SLCentral]

I'm benchmarking for a SLCentral.com review.
Why wouldn't I put everything to high performence? Isn't that one of the benefits to having a good video card? I am playing Unreal 2 at 1280x1024 with high on everything, and it's great. Why would I not deserve the card if I cranked everything to high performence. That makes absolutely no sense.

Guru: Luckily, as I said before, the review is focusing on AA and AF and how it affects the card when gaming/benchmarking. I think'll be good to mention the fact that my processor is not the latest out there, but it is around what most gamers have, making it a good article for the average gamer to read, since most people don't have a 3000+ or a 3GHz processor.

Thanks for everyone's help! I'm benchmarking as I type (I'm using my laptop), and I should be done by the end of the day, so I can work on the actual writing part tomorrow through the rest of the week. Just to let you all know, I'm using the following games/tests for the review (1024x768, 1280x1024, 1600x1200, all tests done with no AA, 2x AA 8x AF, and 4X AA 16X AF)

Quake 3
3DMark2001
3DMark03
Return to Castle Wolfenstein
Unreal Tournament 2003
Comanche 4
Serious Sam 2

If anyone has any other suggestions on games/tests to use for benchmarking, let me know. I want to do more tests, but I'm not too sure about what other tests to do.

 

[EdipisReks]

Originally posted by: SLCentral
I'm benchmarking for a SLCentral.com review.
Why wouldn't I put everything to high performence? Isn't that one of the benefits to having a good video card?

no, the benefits of a good video card are putting everything to high quality, not high performance.

 

[kylebisme]

note that i was responding to Socio in my above post.

also SLCentral, no offence but you realy seem like you should spend more time reading reviews than trying to write them at this point.

Why wouldn't I put everything to high performence? Isn't that one of the benefits to having a good video card? I am playing Unreal 2 at 1280x1024 with high on everything, and it's great.

1) those would be quality settings, something which is had at the expence of prefromace.


2)at those settings, in that game, with your hardware, it is highly unlikely that you are able to sustain a framerate of much over 20fps, it would not shock me if it hit less. it may look realy prety and impressive to you but there is nothing great about a game being rushed out the door with no reguard for industry standers in preformace.

 

[SLCentral]

Although the game was rushed, the game is fun, and performs well on my computer, despite the frame rate being around 25. The game is very smooth, and has no problems graphic-wise.

My mistake about the performance/quality thing, wasn't thinking correctly.

 

[Glitchny]

Originally posted by: Socio

Originally posted by: Viper96720
I meant it's faster than a 4600 with anti aliasing and anistropic filtering or without in certain benchmarks. Some benchmarks are cpu limited so depending on the resolution you won't notice much of a difference. Need to crank up the resolution to really see a difference.

not enabled
enabled

Those links prove the 9700Pro IS NOT FASTER, it just does not slow down as much with everything on.

The fact is with everything cranked to high performance, in all benchmark tests I have run so far the 9700Pro is slighly slower than my old 4600ti.

it is a fact that the 9700pro is faster, its that simple and not only a little faster than a geforce4 its hella faster. its either a cpu limited problem or something else, but there is definatly a large difference between the 4600 and the 9700pro, cause i switched over and the ti4600 cant hold a candle to the 9700pro

 

[SLCentral]

Actually, my specs of my computer are very similar to yours (but you have 1GB of RAM, while I have 512). Although you see a big difference between the two cards, I see almost none.