Need help with this insanity...

[sabbath1]

So I got that P4 2.8C for about two weeks ago along with an ASUS P4P800. In the closest days I got some issues with the board and decided to switch it to a MSI 875P Neo. So I did. Unfortunately, since I had some spare money because of my new job, I decided to do the crazy switch and trade away my brand new 2.8C to a P4 3.0...I did that too.

Now today, when I finally install the not-so-cheap 3.0, and fire up UT2003, I get the same performance as I did with the 2.8! My graphics card is a Radeon 9800 Pro 128 mb, I was benchmarking with the benchmark.exe file that came with the game. I use 6x FSAA and 16x Anisotropic filtering.

My results are for 1280x960: Flyby: 79, Botmatch: 34, and for 1600x1200: Flyby: 74, Botmatch: 24.

Now I'm ashamed to even show this great results. But anyways, does anyone have a clue on why the performance could be the same in UT2003 and seemingly in SOF2 as well?


My system: Intel Pentium 4 3.0, 800FSB / TwinMOS 1Gb DDR400 / MSI 875P Neo / Sapphire Radeon 9800 Pro / Antec True Power 550W / Western Digital Raptor / Hercules G.T XP /
Samsung DVD / Lite-On CDRW

 

[TheMASK]

do u read the reviews at all??? the difference in performance between the springdale and the canterwood is marginal. so is the performance between the 3GHz and 2.8GHz. Of course there shud be a gain with ur new setup, but i dont think it wud be so drastic for you to notice in a game...or, so i think

<b><font color=red>The statement below is True.</font color=red>
<A HREF="http://service.futuremark.com/compare?2k3=959979" target="_new">3DMark 03 score - 297 </A>
<font color=blue> The statement above is false.</font color=blue></b>

 

[sabbath1]

I don't care about performance between S-dale and C-wood, I just knew that poor ASUS board crashed on me, so I rather have a C-wood so I could know for sure I could use PAT. And I might get ECC later anyways, for what it's worth.

Of course, there's no big difference between the 2.8 and 3.0, but at least there should be a difference in the benchmarks like there has always been. Stupid as I was, I only benched like 2 or 3 games with the 2.8 so now I don't have much to compare between.


My system: Intel Pentium 4 3.0, 800FSB / TwinMOS 1Gb DDR400 / MSI 875P Neo / Sapphire Radeon 9800 Pro / Antec True Power 550W / Western Digital Raptor / Hercules G.T XP /
Samsung DVD / Lite-On CDRW

 

[markgun]

I don't think any human would be capable of telling the difference between a 2.8C and 3.0C with a R9800PRO in Unreal2003 . I wouldn't worry about it!

 

[sabbath1]

*sigh* Since I've spent almost $200 more on the 3.0, I sure would want to see a difference. What I meant was that in UT2003, the benchmark told me there was no difference in performance at all, since I got the very same results.

My system: Intel Pentium 4 3.0, 800FSB / TwinMOS 1Gb DDR400 / MSI 875P Neo / Sapphire Radeon 9800 Pro / Antec True Power 550W / Western Digital Raptor / Hercules G.T XP /
Samsung DVD / Lite-On CDRW

 

[Guest]

Well the only thing you have with th 3.0c is better relsell value when PV Prescott comes in. I suggest you sale this processor while its still expensive to buy a newer 3.4 or something...

 

[eden]

Perhaps, just PERHAPS, Windows did not properly adjust to the new CPU.
But I am baffled about this. You are using extremely straining settings to stress the components, and here is where a performance gap can be clearly noticed, as you have bottlenecks to remove.
Now if you wondered visually, of course a 3GHZ would have little difference in FPS during gameplay, however statistically it should yeild about 7% more FPS.

--
I am my own competition. -VJK

 

[sabbath1]

Yeah, before with each graphics card or CPU upgrade, I've noticed difference in games, there really *should* be a difference here too, also, the pack dates of the two cpus only differs by one day, so it can't be a stepping difference, I suppose.

This is really strange...of course, synthetic benches like Sandra shows a difference, but it UT2003 or SOF2 there's none.


My system: Intel Pentium 4 3.0, 800FSB / TwinMOS 1Gb DDR400 / MSI 875P Neo / Sapphire Radeon 9800 Pro / Antec True Power 550W / Western Digital Raptor / Hercules G.T XP /
Samsung DVD / Lite-On CDRW

 

[eden]

They might be locking FPS or using Vsync behind.

--
I am my own competition. -VJK

 

[sabbath1]

Nah, that's not it. Besides, this has lots to do with the graphics drivers. For some reason I always get 34 FPS at botmatch when using 1280x960 res. But at 2x/4x FSAA/aniso or 4x/8x, at 1600x1200, I get over 50 in botmatch! So a part of the problem, or at best, the entire problem lies in limitations, unoptimizations, thus bottlenecks in the Catalysts.

My system: Intel Pentium 4 3.0, 800FSB / TwinMOS 1Gb DDR400 / MSI 875P Neo / Sapphire Radeon 9800 Pro / Antec True Power 550W / Western Digital Raptor / Hercules G.T XP /
Samsung DVD / Lite-On CDRW

 

[eden]

Ahh well, yeah, there's the issue of some bugs in the Catalysts. There is always a special setting that may be buggy. Try at 1024*768.

Additionally it could be an AGP issue on the motherboard's BIOS. Do you have the latest?

--
I am my own competition. -VJK

 

[sabbath1]

Now, some of the times I run the benchmark, I get entirely different results. Seems like fiddling with the control panel creates the difference, even though the settings are the same. Now ATI needs to look at this...

I have the latest BIOS. Unfortunately I must now say this board ain't too stable either. Sometimes it crashes at startup when I've enabled PAT. Could you believe that? I really hope they fix this problem soon. After all, I switched to Canterwood to get rid of this crash issue. I'm in bad luck it seems


My system: Intel Pentium 4 3.0, 800FSB / TwinMOS 1Gb DDR400 / MSI 875P Neo / Sapphire Radeon 9800 Pro / Antec True Power 550W / Western Digital Raptor / Hercules G.T XP /
Samsung DVD / Lite-On CDRW

 

[dwellman]

I guess trying other benchmarks is, now, out of the question since you don't have the Asus anymore. Because, it could be just <i>that</i> particular benchmark, right?

But really, I have nothing to add to what Eden has already suggested.

As far as ATI goes, I hope you have better luck than I did. Their tech support is abyssmal. (Me: <i>I have problem X. I've done A, B, and C.</i> ATI: <i>Thank you for choosing ATI. Many issues can be solved by doing A, B, and C.</i>) One can go around in circles for weeks.

Dave

__________________________________________________
<b><font color=red>Three great virtues of a programmer are: laziness, impatience, and hubris.</font color=red><b>

 

[sabbath1]

Actually after a few reruns, I got a few extra FPS out of the P4 3.0, especially in benches without AA/AF, or with "just" 2x/4x. So I'm sure I'm now enjoying the slight but existent increase.

My system: Intel Pentium 4 3.0, 800FSB / TwinMOS 1Gb DDR400 / MSI 875P Neo / Sapphire Radeon 9800 Pro / Antec True Power 550W / Western Digital Raptor / Hercules G.T XP /
Samsung DVD / Lite-On CDRW

 

[JP5]

It probably wasnt worth it to spend $200 more to swap out the 2.8 for 3.0 But don't worry about it at this point -- you've still got one of the fastest comps on the block. Especially after you bump up to 3.5

 

[sabbath1]

Actually I only bought the 3.0 because of my new job. Otherwise, I'd kept my 2.8. The sour stuff was that I got the job just one day after I took out my 2.8, and thus it was already used. But I couldn't resist reaching the 3 Ghz mark, so thus the order was born.

My system: Intel Pentium 4 3.0, 800FSB / TwinMOS 1Gb DDR400 / MSI 875P Neo / Sapphire Radeon 9800 Pro / Antec True Power 550W / Western Digital Raptor / Hercules G.T XP /
Samsung DVD / Lite-On CDRW