Got an X1900XT - Does my 3dmark05 score indicate a problem?

[Cadaver]

Running with the system in my sig, I got 8958 in 3dmark05. Does it indicate anything untoward (apart from the fact that I am CPU bottlenecked)?

 

[aggressor]

Sounds about right. I got 10527 with a X1900XT and a A64 3200+ @ 2.5ghz.

 

[PWMK2]

You're not CPU bottlenecked lol...

 

[Cadaver]

Sorry. Bottlenecked is the wrong word. I just meant to explain why the score was lower than that of other systems with X1900XTs,which frequently score substantially higher.

Thanks for the advice. Now I know all is well.

 

[Rhitick]

Yeah its the proc, mine scores that pre-OC. 2.5ghz yields 10627 or somewhere around there...it's in my sig.

But otherwise your fine. If you raise AA and AF the score might even stay the same or close.

 

[Eastcoasthandle]

You're not CPU bottlenecked lol...

That is correct the CPU is a bottle neck for these cards. In order to get better performance you need a fast dual core solution....

 

[PWMK2]

That is correct the CPU is a bottle neck for these cards. In order to get better performance you need a fast dual core solution....

LOL, no you don't. Well, in 3DMark06 dual-core processors score higher but in reality there is little to no performance difference in actual games. It's not bottlenecking. Some people don't realize what bottlenecking really is. Basically, you'll need a Sempron or something to bottleneck that card...

 

[Eastcoasthandle]

3dmark05/06 is CPU limited bottom line. There is only 1 bench that I know of and GPU limited the rest is depended on how well your clock your CPU, bottom line.

If you don't understand then do this.
-run 05/06 at your normal cpu/gpu clock
-next oc your cpu as high as you can.

Is the score the same? If this is gpu specific benchmark then you are suppose to get the roughly the same score until you OC your GPU. (+/-) a few 3d Marks of course.

 

[Eastcoasthandle]

Here let me answer my very own question for you by showing you a few pics

This pic is of top Intel P4 scores using a x1900xtx

Next, is top scores using A64 again with x1900xtx

Finally with A64 X2

Did you notice a trend? Well let me spell it out for you...the better the processor the better the benchmark. No matter how well you OC your VC, it's the processor that ultimately increase benchmark as well as frame rate in games. As I was saying in my previous post the GPU compliments the CPU, not the other way around. Games are very dependant on CPU moreso then GPU. Regardless if its Fear or Q4 you still need a good CPU to play them properly.

 

[PWMK2]

Uh... you know what... I never said that a better CPU won't increase your scores. However, what you need to realize is that all of the pics you just showed were setups where the CPU was overclocked up the wazoo. My 3dMark05 score increases a measly 300 points going from 3.0GHz to 3.4GHz, which is a typical overclock. (From about 6100 to 6400). Games are rendered on GPUs, not CPUs, although a faster CPU speed will speed up things up because the CPU is still stuck handling all of the processes.

Saying games are dependent on the CPU is idiotic. A better CPU will allow games to play better, yes, but have you noticed how it's always the better GPU that wins in the long run? Whenever people upgrade for gaming, they are always advised to upgrade their video cards first, not their processors, and with good reason - you're not going to get the same performance increase going from even an X1800XT to an X1900XTX as you would going from simply a 3000+ to something like a 3700+. It just won't happen. If you could choose to get an OK processor with an awesome video card, or an awesome processor with an OK video card, what would you get? Basically, would you choose the system with a 3200+ and an X1900XTX or the system with the FX-60 and the X1800XL? Well... I know which system would play games better, and which one I would use (hint: it ain't the FX-60 based system).

 

[Eastcoasthandle]

Uh... you know what... I never said that a better CPU won't increase your scores. However, what you need to realize is that all of the pics you just showed were setups where the CPU was overclocked up the wazoo. My 3dMark05 score increases a measly 300 points going from 3.0GHz to 3.4GHz, which is a typical overclock. (From about 6100 to 6400). Games are rendered on GPUs, not CPUs, although a faster CPU speed will speed up things up because the CPU is still stuck handling all of the processes.

Saying games are dependent on the CPU is idiotic. A better CPU will allow games to play better, yes, but have you noticed how it's always the better GPU that wins in the long run? Whenever people upgrade for gaming, they are always advised to upgrade their video cards first, not their processors, and with good reason - you're not going to get the same performance increase going from even an X1800XT to an X1900XTX as you would going from simply a 3000+ to something like a 3700+. It just won't happen. If you could choose to get an OK processor with an awesome video card, or an awesome processor with an OK video card, what would you get? Basically, would you choose the system with a 3200+ and an X1900XTX or the system with the FX-60 and the X1800XL? Well... I know which system would play games better, and which one I would use (hint: it ain't the FX-60 based system).

Tell you what take your theory and try it out for yourself for a change and post some facts. All you doing is spitting out opinions that are not true.

Look here
for the Fear benchmark. Both Conroe and FX-60 are using X1900xt in CF mode clocked the same using Maximum graphics quality setting @1280x960 (second bench). As you can see the Conroe takes a substantial lead over the FX-60 and the VC are not even OC. Therefore, to say that ..." games are dependent on the CPU is idiotic" is idiotic in itself, an oxymoron of sorts. All I can do is link you to facts. So by all means post your opinion but they are not facts. By using this link it clearly shows a substantial increase in min, average and max frame rates (in this case FPS). This is not defined as "better game play" as you put it but an overall increase in performance. Better game play in my book is a mere increase in FPS but as you see here even the min. fps was increase by the Conroe (in this case 5 FPS 65 vs 60 FPS). As your arguement suggest this aspect is to be by the GPU only. However, in this benchmark it is found to be false (a good CPU can also do this as well). Even if we used the max FPS, it's the Conroe complimented with x1900xt CF that gave the best performance. However it was the Conroe alone that showed the increase vs Fx-60 as the x1900xt was not OC. Therefore, you are again incorrect.

Therefore, I will say it again, the GPU compliments the CPU in gaming not the other way around. There is no middle ground you do need a good CPU and GPU however, a GPU that is better then CPU will only create a bottleneck.

 

[BBA]

3dmark05/06 is CPU limited bottom line. There is only 1 bench that I know of and GPU limited the rest is depended on how well your clock your CPU, bottom line.

Your losing credibility day after day here man...you need to stop acting like you know it all. The 3DMark06 does improve slightly with CPU increases, but trust me, it is FAR from being CPU limited. A faster video card makes MUCH more difference than a faster CPU. Not exactly the definition of CPU limited, unless you want to run it at like 640x480 or something (if thats even possible?)

But...that score is perfectly normal for an untweaked/unoverclocked X1900XTX system on an AMD 2.0 GHz chip.

 

[BBA]

Tell you what take your theory and try it out for yourself for a change and post some facts. All you doing is spitting out opinions that are not true.


... All I can do is link you to facts. So by all means post your opinion but they are not facts.

Therefore, I will say it again, the GPU compliments the CPU in gaming not the other way around. There is no middle ground you do need a good CPU and GPU however, a GPU that is better then CPU will only create a bottleneck.

Then why does a 3200 perform within a couple of FPS of an FX57 in Quake4?

I think your just being taken in by benchmarks made to push the superiority of the new 'intel' CPU.

You need to live in real life here man...no more marketing, your swayed too easily.

.
Oh...you wanted to see proof...HERE:

http://www.xbitlabs.com/articles/cpu/display/cpu-games2_6.html

 

[Eastcoasthandle]

Then why does a 3200 perform within a couple of FPS of an FX57 in Quake4?

I think your just being taken in by benchmarks made to push the superiority of the new 'intel' CPU.

You need to live in real life here man...no more marketing, your swayed too easily.

.
Oh...you wanted to see proof...HERE:

http://www.xbitlabs.com/articles/cpu/display/cpu-games2_6.html

That only proves my point. You are merely twisting words. But for the record lets repeat them again.

Therefore, I will say it again, the GPU compliments the CPU in gaming not the other way around. There is no middle ground you do need a good CPU and GPU however, a GPU that is better then CPU will only create a bottleneck.

That is what you quoted. Be that as it may that is single core vs dual core. Which wasn't my my point at all. Again read the above for details.

Therefore, I will say it again, the GPU compliments the CPU in gaming not the other way around. There is no middle ground you do need a good CPU and GPU however, a GPU that is better then CPU will only create a bottleneck.

Here I quoted it again for clarity.

 

[Tetrahedron]

I get 11512 in 05 and 5919 in 06

 

[Eastcoasthandle]

I get 11512 in 05 and 5919 in 06

Impressive using an Opty 1900xt combo.

 

[Tetrahedron]

Impressive using an Opty 1900xt combo.

really? Is there a problem with Opty's and ATi or just the X1900?

 

[Jodiuh]

I'm at 12480 in 05.

 

[PWMK2]

Tell you what take your theory and try it out for yourself for a change and post some facts. All you doing is spitting out opinions that are not true.

Well, since you like "posts for clarity" so much, how about actually reading what I posted:

My 3dMark05 score increases a measly 300 points going from 3.0GHz to 3.4GHz, which is a typical overclock. (From about 6100 to 6400).

I don't really want to have to redo the scores again (I deleted those scores in my ORB) but I will if need to.

Truth be told, yes, a better cpu will result in better gaming. But games will run like crap if you have a crappy video card, not necessarily if you have a crappy processor. Comparing an FX-60 to Conroe using low-quality settings doesn't work anyway... in order to get those kind of scores in F.E.A.R. that you linked to with any modern video card, you're pretty much having to set the resolution to 640x480. Not exactly the best way to benchmark CPUs under normal gaming circumstances.

really? Is there a problem with Opty's and ATi or just the X1900?

No, at least none that I have heard of.

 

[Jodiuh]

Oh I beg to differ. The problem with Opty + 1900 is the "extended belly" region due to lack of "away from computer" time.

 

[BBA]

Impressive using an Opty 1900xt combo.

Why do you say that? I say it's normal.

My Stock Opty 175 (2.2) gets 10890 3dmark05 and 5990 3dmark06 with my X1900xtx.

Now..on to your argument about CPU making the difference...it's the combination of the whole system, cpu, drivers, vid card, mem, etc... it just so happens that the vid card makes the BIGGEST difference in performance of ANY of the rest (well, a hosed OS can kill a score pretty badly too).

We all know the CPU makes only a minor difference...you have to do some really impressive cpu change (like adding a hardware physics processor or something) to make a big difference in video performance.