• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

*Updated* 3DMark scores dropped just by switching the Mobo?

J

jtusa

Diamond Member
Was having stability problems with my Intel D865GBF, so I switched to a Asus P4P800 Deluxe. It's a very stable system now, however, graphic performance has suffered.

3DMark03 scores went from 5600ish to 4500ish and 3DMark05 scores went from 900ish to 500ish. All hardware switched over is the exact same from before, the only difference is the mobo(and a new HSF, but that doesn't really matter here).

Using same Nvidia drivers as before(the latest, 61.77 I think it is), plus latest updates/drivers for the mobo.

Specs:

P4 3.0G/512K @ 3.3Ghz
P4P800 Deluxe
1024MB PC2700 in dual channel
Nvidia 5900 128MB

Ideas? 😕

Update:
Today I did some testing. First, per the general consensus, I removed the overclocking since the memory wasn't letting me really take advantage of it. Then I ran the following tests:

61.77 drivers, Memory in dual-channel
3DMark03 = 4670
3DMark05 = 566

66.93 drivers, Memory in dual channel
3DMark03 = 5231
3DMark05 = 664

66.93 drivers, Memory not in dual-channel
3DMark03 = 5132
3DMark05 = 667

So the drivers made a noticeable difference, but it seems running memory in dual channel or not doesn't seem to matter. However, average scores are still lower than they were with my Intel board(5600ish in '03 and 900ish in '05).
 
Ya there is something real wrong with your scores period. here;s mine
3Dmark 03 13953 cpu@3561 2-5-2-2 Ballistic Ram@222
3Dmark 05 6176 cpu@3523 2-5-2-2 Ballistic Ram @220
PCmark 04 5700 cpu @3590 2-5-2-2 BH-5 @224
Back in bit there not publishing My 3Dmark 03 score pisses me off
YOUR video card is not working correctly if those are infact your scores which I find hard to believe
 
why are you using pc2700 ddr with an 800mhz fsb cpu? you are starving your cpu. Try the beta nvidia drivers and see if they help your score. Did you re-install windows after switching the motherboard and ensure you installed all the chipset drivers? There could also be some bios tweaking you could do to get your score back up to where it should be.
 
Microsoft Windows XP Microsoft Windows XP

DirectX Version 9.0c 9.0c

Mobo Manufacturer http://www.abit.com.tw/ http://www.abit.com.tw/

Mobo Model IC7-MAX3(Intel i875P-ICH5) KV8 Pro (VIA K8T800P-8237)

AGP Rates (Current/Available) 8x / 4x, 8x 8x / 4x, 8x

CPU Intel Pentium 4 3561 MHz AMD Athlon(tm) 64 2401 MHz

FSB 222 MHz 204 MHz

Memory 1024 MB 1024 MB



Display Information
Graphics Chipset ATI RADEON X800 XT PE ATI RADEON X800 XT PE

Driver Name RADEON X800 XT Platinum Edition RADEON X800 XT Platinum Edition

Driver Version 6.14.10.6467 6.14.10.6490

Driver Status WHQL - FM Approved WHQL - FM Approved

Video Memory 256 MB 256 MB

Core Clock 547 MHz 540 MHz (mine)

Memory Clock 609 MHz 608 MHz ( AMD64)



Sound Information
Sound Adapter Driver Name Realtek AC97 Audio SB Audigy 2 ZS Audio [B000]

Sound Adapter Driver Version 5.10.0.5350 5.12.6.402



Benchmark Settings
Program Version 3DMark03 Revision 4 Build 0 3DMark03 Revision 5 Build 0

Resolution 1024x768@32 bit 1024x768@32 bit

Texture Filtering Optimal Optimal

Pixel Processing / Antialiasing None None

Post Processing false false

Vertex Shaders Optimal Optimal



Main Test Results
3DMark Score
(3DMarks) 13953 ( MY score)

13336 (amd64)

I have the 12 highest score in this bench
 
Those are in fact my scores. I did a fresh install of XP after switching out the mobo. I'm using the 2700 because I took it from older system and just haven't gotten around to shelling out the cash for a new dual channel pair.

I'll try the beta drivers and see how much of a difference that makes.
 
Information
Operating System Microsoft Windows XP Microsoft Windows XP

DirectX Version 9.0c 9.0c

Mobo Manufacturer http://www.abit.com.tw/

Mobo Model IC7-MAX3(Intel i875P-ICH5) MS-7025

AGP Rates (Current/Available) 8x / 4x, 8x 0x /

CPU Intel Pentium 4 3523 MHz AMD Athlon(tm) 64 3572 MHz

FSB 220 MHz 200 MHz

Memory 1024 MB 512 MB



Display Information
Graphics Chipset ATI RADEON X800 XT PE ATI RADEON X800 XT PE

Driver Name RADEON X800 XT Platinum Edition (Omega 2.5.90) RADEON X800 XT Platinum Edition

Driver Version 6.14.10.6490 6.14.10.6490

Driver Status Non WHQL - Not FM Approved Non WHQL - Not FM Approved

Video Memory 256 MB 256 MB

Core Clock 550 MHz 823 MHz

Memory Clock 609 MHz 658 MHz



Benchmark Settings
Program Version 3DMark05 Revision 1 Build 0 3DMark05 Revision 1 Build 0

Resolution 1024x768@32 bit 1024x768@32 bit

Anti-Aliasing None None

Texture Filtering Optimal Optimal

Vertex Shader Profile 2_0 2_0

Pixel Shader Profile 2_b 2_b

Force Full Precision Off Off

Disable DST Off Off

Disable Post-processing Off Off

Force Software Vertex Shaders Off Off

Color Mipmaps Off Off

Repeat Count Off Off



Main Test Results
3DMark Score
(3DMarks) 6176 (my score)

8532 (AMD the top score)







 
General Information
Operating System Microsoft Windows XP Microsoft Windows XP

DirectX Version 9.0c 9.0c

Mobo Manufacturer http://www.abit.com.tw/ INTEL

Mobo Model IC7-MAX3(Intel i875P-ICH5) CANTERWOOD

AGP Rates (Current/Available) 8x / 4x, 8x 8x / 4x, 8x

CPU Intel Pentium 4 3590 MHz Intel Pentium 4 3986 MHz

Physical CPUs 1 2

HyperThreading Available - 2 Logical Processors Available - Disabled

FSB 224 MHz 532 MHz

Memory 1024 MB 1022 MB



Display Information
Graphics Chipset ATI RADEON X800 XT PE NVIDIA GeForce 6800 Ultra

Driver Name RADEON X800 XT Platinum Edition NVIDIA GeForce 6800 Ultra

Driver Version 6.14.10.6467 6.6.8.1

Video Memory 256 MB 256 MB

Core Clock 540 MHz 5 MHz

Memory Clock 608 MHz 6 MHz



Sound Information
Sound Adapter Driver Name Realtek AC97 Audio

Sound Adapter Driver Version 5.10.0.5350



Hard Disk Information
Model

Connection

Capacity 0.0 GB 0.0 GB



Benchmark Information
Program Version PCMark04 Revision 2 Build 0 PCMark04 Revision 2 Build 0



Main Test Results
System Test Suite
(PCMarks) 5700 (my score)

8092 (the top score)


 
EVEN if you only have PC2700 those scores are way off if i were to guess your graphics card is not write
E- mail me at (raidcomputers@msn.com) and I will give you the optamization set up for gaming it will take you a little while to set up but u need help I well e-mail you back with all the information you need to get it running way better than what I am seeing ALSO could you go to the orb and do the same thing I did just copy post so i can have a look or if your score are published what is your header that would be the best thing. than i can go check it out ( do not install the Beta drivers)
 
Zinn, he is using an Nvidia 5900, you are using an x800, of course his scores are going to be lower than yours. And the scores he was originaly getting are just about right for his card. The reason I told him to go with the beta drivers, is because with the 61.77 drivers with my 6800gt, it greatly reduced my score, and also cause lagging/stuttering in games. After going to the beta drivers, my scores went up to where they were supposed to be, and the lagging/stuttering went away.
 
ya i didn't expect his scores to be like mine but they should be close to my 9700 pro scores there know were near them 03 8800 9700 pro )05 4200 700 pro
 
Aside from some great advice already given in previous replies (drivers, OS reload, etc.) you might wanna check to see if the BIOS on your new mobo has any updates for it. Since the only thing changing is the board, it has a pretty high chance of being the culprit. Barring success from that avenue, a total reload is probably the next best choice to try out.
 
Hold your horses, his scores are right. You cannot compare your scores to an x800 as said above. I personally have a 2.4 P4(533 running 166 to 3.0) with a Radeon 9800pro and get around 6900 on 03 and 2400 on 05.

He is using a 5900 (decent card) but not in dual channel so theres the score drop. Just one minor point, if you are o/c your system, don't foget it increases demand on the memory as well as the chip so 'can' actually have the reverse effect. try putting your memory in slots 1 and 3 to stop the dual channel kicking in see if that makes a difference, seeing your memory isn't rated for an 800 bus never mind a 880 internal clock. Your memory is 333 hz rated, which in theory equates to the 664bus mark(i think). To o/c an 800 bus to 880 you really should have ddr500.

Try running your pc at stock, memory @ spd and see what you get from there. Then o/c and check results. More aggressive memory timing sometimes give better results than o/c of a cpu, the 10 % o/c you are doing is hardly worth killing your pc for, especially as your memory cannot hadle it. With your memory in asynchronous mode, 333, try ramping the memory timing to 2T 2,2,6 or maybe even go for 5. 2.7 volts to the ram will keep it stable. your saving to get some more memory and 2.7 volts will take ow 4 years to burn out most memory. You'll have another pc by then so no worries.

Run sandra to check your memory b/w and you'll be suprised at the difference.
 
Are you suggesting that your score in 3d mark 05 should better than his score. I think not It says in his post his running at PC 2700 ddr in dual channel A score of 900 in 3D05 give me a break.What you don't think I am streesing my memory running at 1to1 @222 AT 2-5-2-2 pat enable lets be real hear and he is not running at 666FSB its closer to 800 FSB he is O/C to 3.3 Ghz how do you think he got there .ALSO he is probably running insyn. just as I am.His 3d05 score should be between 4200 to 4500 range in3D o3 his score should be in the 8000to 9000 range 1 more thing my memory is in slots 1and 3 just as his should be.
Also i gave him my scores with a 9700pro installed its in the post that also had A 2.8C GHz installed .ONE more think I don't buy memory unless it has a lifetime warranty
 
HE IS RUNNING A GEFORCE FX IDIOT!!!

3dMark03 should be ~5200
3dMark05 should be ~1300 with approved drivers.

He WILL NOT get 9000 on 03 with any card lower than an 6800NU. 9800Pros only make ~6000.

Zinn you are FULL OF CRAP:

STFU Troll

-Kevin
 
I'm betting because your older motherboard defaulted on a different divider than the newer mobo resulting from the pc2700 ram. The newer motherboard is likely stable because its working on a divider or auto configuration safe enough to run at that clock speed without burning up the ram.

Get PC3200 or Pc3500 (if you want to overclock).
 
Again gamegeek Haven't I embarrassed you enough all ready I said between 8000to 9000 AND I got 8800 with a 9700pro fool With a 2.8C P4 you really are stupid. By THE way read the latest post
 
I went to a LAN party once, meeting up with guys I'd played games with for years. One of them boasted about how he got 'amazing' scores with 3D Mark '03 on his ramped up super charged system. Guess what, turns out he was full of crap too. No such machine with him, brought his 'sisters' instead

Now these things are FACT not fiction.

1. 3D Mark '03 is buggy, its artificial, and anyone who gets the exact same results from one day to the next is as above, full of crap.
2. Radeons blow the floor with the Geforce FX5800/5900 rangeon 3D mark '03, irrespective of your FSB. This is due to point 1, not due to the Radeon being better, its just that futuremark developed '03 with money and testing rigs from ati, Nvidia refused to give them any for development.
3. As my post said earlier, I too had a 9700pro, even with my system running at 3.0ghz, and the 9700 pushing it a little, wouldn't get much over 5000. That was the industry joke at the time as 3 months after it hit the shelves they released the 9800pro which hit 6000 easy, no o/c.
4. it is impolite to reply to a post without actually reading it.
5. With or without dual channel, systems run far smoother when memory is in synchronous mode. More of a performance gain is shown in sync rather than async. The 664 bus speed, which actually is the 4 times multipul of pc2700 memory is its top rated in sync memory timing so an in sync bus of 880 will not work.
6.As I said in my reply, I know exactly what bus speed he is at hence I said 880 internal clock. Through (thru to our colonial bretheren)the genius of mathematics (math to our colonial....) I worked it out.
7. Lifetime waranties for memory are void when burned out due to excessive voltage applied.
8. whether it be a multipul of 400, 533, 800, a 3 ghz clock speed is just that, 3 ghz. The bottleneck of the system bus will slow down to the slowest part of your system, which in gfx cases is your AGP bus. So whether you are using a 666 bus or an 800 bus will not matter in gfx scores that much, its the raw speed that counts.

lesson finished.

His 3D '03 score should be around the 5500-mark, due mostly to his processor/gfx combination if he was using in sync memory and as I said, its a decent gfx card( bet its plays games better than my stupid arsed radeon crash bollix drivers bollix every month bollix new game finds new crash bollix.....yes i hate my radeon, but best bang / buck as they say).

My point was to show there was more than one reason his scores were way off.


Thanks Kevin for that brilliant broadside...pmsl at that one. Thx also to Reg who obviously read my post( see 4 above very polite man).
Just to throw spanner in works on Reg's explanation, I thought P4 dual channel only worked in synchronous mode, and Nforce chipped AMD's where the only ones that could run dual channel async? That is why I suggested to put memory in slots 1 and 3 to stop it happening anyway.On P4 board's Zinn2b, putting memory in slots 1 and 3 STOPS dual channel, they need to be matching pairs in side by side slots. Could of sworn its the same on AMD platforms as well.

yours form Blighty and off to play some COD,

Bill
 
Originally posted by: Zinn2b
Again gamegeek Haven't I embarrassed you enough all ready I said between 8000to 9000 AND I got 8800 with a 9700pro fool With a 2.8C P4 you really are stupid. By THE way read the latest post
Click to expand...

This post isn't a means for you to showoff your hardware and benchmarks. If you're not going to help then just don't post. You've done nothing but prove to everyone, including Gamegeek that the only embarrassment going on is what you're bringing on yourself by what you've replied with here. I've read 20 posts probably by you today and every one of them is as insightful as your post about needing a 480w PSU to run a 6800. If you really do have all the hardware you've claimed to have owned on this board in the last 24 hours, you've done nothing here but prove people with nice things can still be unbelievably ignorant, misinformed and gullible.



As to the actual topic at hand, the driver issue would be my first guess as well, since I noticed the same drop and increase with the beta drivers. I also agree that 3dmark03 is severely touchy. Sometimes I'll completely reboot, dump the memory, close all programs and run 3dmark03 only to wind up with a terribly lower score than when I had 20 programs running, 1/2 the RAM available and a system that's been running for 2 days. I wouldn't put too much worry into your scores unless you see an actual degrade in real world gaming performance, but again try the driver upgrade. It helped me a bit.






 
Originally posted by: Zinn2b
Again gamegeek Haven't I embarrassed you enough all ready I said between 8000to 9000 AND I got 8800 with a 9700pro fool With a 2.8C P4 you really are stupid. By THE way read the latest post
Click to expand...

His score WILL NOT be around 8000-9000. What DO YOU NOT understand about this. THERE IS NO WAY a 9700Pro will score that high either. The MINIMUM card needed to get that i would say is a 6800NU.

Its gamingphreek dumbass.

Im really feeling embarassed :roll: If you haven't noticed no one likes you and it would be better to just leave. It would also be better if you:

STFU TROLL

-Kevin
 
Actually, from the comments he made over in another thread around 11am this morning, it looks like Zinn2b got slapped with a permanent STFU from the mods:

zinn2b: ADUL are you a Moslim? Got any bombs? I think maybe you should be checked out!!!

---

Zinn2b,

Do you ever post anything but personal insults? I think you should check out.

Good bye.

AnandTech Moderator
Click to expand...

Hopefully, we won't be hearing any more from ol' Zinn.
 
Originally posted by: batmanuel
Actually, from the comments he made over in another thread around 11am this morning, it looks like Zinn2b got slapped with a permanent STFU from the mods:

zinn2b: ADUL are you a Moslim? Got any bombs? I think maybe you should be checked out!!!

---

Zinn2b,

Do you ever post anything but personal insults? I think you should check out.

Good bye.

AnandTech Moderator
Click to expand...

Hopefully, we won't be hearing any more from ol' Zinn.
Click to expand...

:beer:
 
So the drivers made a noticeable difference, but it seems running memory in dual channel or not doesn't seem to matter. However, average scores are still lower than they were with my Intel board(5600ish in '03 and 900ish in '05).
Click to expand...

Well different boards can be better/worse with 3d performance, but it's usually a small margin. I think you've nailed your biggest problem though. The overclock could've easily made up the difference in your score now. Congrats on the solution 😛

As for Zinn that's great to hear.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top