Is my 3DMark03 score on par with the system?

[Trente]

AMD Athlon 64 3200+ 'Winchester'

Gigabyte K8NF-9 nForce4 MCPs

1GB Samsung Original 400MHz (Dual-Channel)

Gigabyte GeForce 6600GT 128MB PCI-e

WD 160GB w/8MB 7200RPM S.ATA

Philips Acoustic Edge 5.1 PCI


Nothing overclocked, running default settings.

Result: 7896

 

[imverygifted]

sounds good, my agp 6800gt gets about 11000 stock and 12500 ultra that sounds like the difference between the 2 cards

 

[mwmorph]

yeah. uyour 6600gt is holding the system back. you could get at least 10K with a 6800nu oced and unlocked.

 

[gamerj]

Well, u could ask it on the 3dmark forum lol...

By the way... try this benchmark: http://www.aquamark3.com

 

[ComatoseDelirium]

Thats actually a tad low, I have the same video card as you and half the ram, and a slower processor, P4 3.0, and I get 8500

 

[Trente]

Originally posted by: ComatoseDelirium
Thats actually a tad low, I have the same video card as you and half the ram, and a slower processor, P4 3.0, and I get 8500

What driver version are you using?

Also, I remember reading 3DMark tends to like P4 over Athlon... Am I right?

Could this be the explanation for my lower result?

 

[DrCool]

yeah, your score is on the low side..

download sisoft sandra, and run the memory bandwidth benchmark.

you should get somewhere in the 6000 range, if your not, you don't have something setup correctly.

i'm getting 8087 with half the memory on a nForce3 board (single channel)

 

[Trente]

http://forums.anandtech.com/messageview.aspx?catid=27&threadid=1508070

Could this have a connection with the low result?

I don't get it, I got a quality 1GB RAM running in Dual-Channel, an NF4 board and a 3200+ Winchester, but my score is lower than you guys?

 

[Glpster]

Originally posted by: DrCool
yeah, your score is on the low side..

download sisoft sandra, and run the memory bandwidth benchmark.

you should get somewhere in the 6000 range, if your not, you don't have something setup correctly.

i'm getting 8087 with half the memory on a nForce3 board (single channel)

6000 Memory Bandwidth?

How do you figure?

The best references in Sandra aren't even that high. And the highest reference is PC5300 DDR2 Memory! (Which is 5704 MB/s)

I'd certainly like to know how you are getting 6000 for your memory bandwidth, because if that's the case then my bandwidth is really low at 5264.

 

[DrCool]

i'm assuming you read anandtech, since your posting on the anandtech forums..

http://images.anandtech.com/graphs/corsair%204400c25amd_01030580108/5942.png

dual channel - stock DDR400 speeds!

 

[Duvie]

Originally posted by: Glpster

Originally posted by: DrCool
yeah, your score is on the low side..

download sisoft sandra, and run the memory bandwidth benchmark.

you should get somewhere in the 6000 range, if your not, you don't have something setup correctly.

i'm getting 8087 with half the memory on a nForce3 board (single channel)

6000 Memory Bandwidth?

How do you figure?

The best references in Sandra aren't even that high. And the highest reference is PC5300 DDR2 Memory! (Which is 5704 MB/s)

I'd certainly like to know how you are getting 6000 for your memory bandwidth, because if that's the case then my bandwidth is really low at 5264.

Ohh how little the newbs Know!!!

7300's

7200's

In both of those cases I wasn't running 1:1 test cause I had a bad stick...RMAing now...I have rasn with no screenies mem scores at high as 7600 but was not stable in testing...I have even seen 8000 with 1:1 and 300fsb....

Do some research the p4 and DDR2 are not the memory bandwidth powerhouses....unfortunately the AMD cant really seem to take adavantage of it...

 

[ComatoseDelirium]

I have the 71.80 ForceWare drivers from guru3d, its odd that mine is higher than you seeing as you have a faster processor, and double the ram, but benchmarks can be wrong.. its very general, if it plays games well and is fast and responsive.. than screw the mark.

 

[Glpster]

Originally posted by: Duvie

Originally posted by: Glpster

Originally posted by: DrCool
yeah, your score is on the low side..

download sisoft sandra, and run the memory bandwidth benchmark.

you should get somewhere in the 6000 range, if your not, you don't have something setup correctly.

i'm getting 8087 with half the memory on a nForce3 board (single channel)

6000 Memory Bandwidth?

How do you figure?

The best references in Sandra aren't even that high. And the highest reference is PC5300 DDR2 Memory! (Which is 5704 MB/s)

I'd certainly like to know how you are getting 6000 for your memory bandwidth, because if that's the case then my bandwidth is really low at 5264.

Ohh how little the newbs Know!!!

7300's

7200's

In both of those cases I wasn't running 1:1 test cause I had a bad stick...RMAing now...I have rasn with no screenies mem scores at high as 7600 but was not stable in testing...I have even seen 8000 with 1:1 and 300fsb....

Do some research the p4 and DDR2 are not the memory bandwidth powerhouses....unfortunately the AMD cant really seem to take adavantage of it...

Yes, but aren't you overclocking there?

I thought the claim was that these bandwidths should be around 6000 at stock speeds. I figured I was doing good, since my readings were fairly high up the reference benchmarks in Sandra, but supposedly that is not the case.

So, I wonder what I'm doing wrong then.

As mentioned in another post, I discovered looking at CPU-Z that my memory timings are set to 3-4-4-8 at 200MHz. Although NewEgg claims these are 2.5-3-3-7 chips at PC3200. PQI Dual Channel Turbo Memory

The SPD ratings in CPU-Z state that 2.5-3-3-7 is for 166 MHz!! So I think I was taken in by some false advertising there.

At any rate, is there something that I (and others) should be doing in order to achieve these higher bandwidths without overclocking?

Your advice would be greatly appreciated.

 

[boyRacer]

Got a question for you... does your Acoustic Edge still work with most of the games you play? Philips abandoned driver support... but I loved that card.

 

[Trente]

Originally posted by: bR
Got a question for you... does your Acoustic Edge still work with most of the games you play? Philips abandoned driver support... but I loved that card.

Yes, it does. I was surprised to discover that it works flawlessly on my nForce4 platform. It's a shame that Philips abandoned driver support; It seems they actually dropped the number of voices the card can produce on the latest and last driver set (v1.89), as it fails the 60 voices test on 3DMark. Do you think it has anything to do with my low score?

Could it be that the Philips uses many CPU cycles that it affects my 3DMark score?

 

[Trente]

I went over some reviews of the 6600GT again, checked 3DMark03 benchmarks... Using the same Gigabyte card as mine, guru3d.com were able to score 8064 points; that's ~150 points more than what my system was able to score, but they used this system for the test:

* Albatron PX915P/G Pro (PCI-Express 16x enabled)
* 1024 MB DDR400
* Pentium 4 class 3.6 Ghz (Socket 775)
* Windows XP Professional
* DirectX 9.0c
* ForceWare 6681 WHQL
* Latest reference chipset and PCI-Express drivers

That CPU is more powerful than mine, they also got the latest P4 chipset as their platform along with 1GB of RAM. So it seems my score is on par after all...?

 

[danteshors]

I scored a 12800 with my P4 setup and 6800 GT at ultra speeds on my first try. I thought the AMD's were supposed to be better at game benchmarks than that.

P4 3.0 GHZ @ 3.3Ghz
2GB RAM
6800GT at 410 and 1110
74GB Raptor is my main drive with a few other 120 GB'ers


Just as a comparison.

 

[Chosonman]

I have a AMD 3000 Newcastle
6800 NU 128 MB
1024 Corsair Value RAM
Western Digital HD

Scored: 11710 on 3DMark03

 

[DusterAZ]

Originally posted by: Trente
I went over some reviews of the 6600GT again, checked 3DMark03 benchmarks... Using the same Gigabyte card as mine, guru3d.com were able to score 8064 points; that's ~150 points more than what my system was able to score, but they used this system for the test:

* Albatron PX915P/G Pro (PCI-Express 16x enabled)
* 1024 MB DDR400
* Pentium 4 class 3.6 Ghz (Socket 775)
* Windows XP Professional
* DirectX 9.0c
* ForceWare 6681 WHQL
* Latest reference chipset and PCI-Express drivers

That CPU is more powerful than mine, they also got the latest P4 chipset as their platform along with 1GB of RAM. So it seems my score is on par after all...?

Yup, I wouldn't stress you seem to be in the ball park. Plus the PCIe version of the 6600GT runs with memory at 1000 Mhz default vs 900 Mhz on the AGP version. I wouldn't be surprised if you were faster than that system by simply overclocking your memory to 1000 Mhz.

 

[Trente]

Originally posted by: DusterAZ

Originally posted by: Trente
I went over some reviews of the 6600GT again, checked 3DMark03 benchmarks... Using the same Gigabyte card as mine, guru3d.com were able to score 8064 points; that's ~150 points more than what my system was able to score, but they used this system for the test:

* Albatron PX915P/G Pro (PCI-Express 16x enabled)
* 1024 MB DDR400
* Pentium 4 class 3.6 Ghz (Socket 775)
* Windows XP Professional
* DirectX 9.0c
* ForceWare 6681 WHQL
* Latest reference chipset and PCI-Express drivers

That CPU is more powerful than mine, they also got the latest P4 chipset as their platform along with 1GB of RAM. So it seems my score is on par after all...?

Yup, I wouldn't stress you seem to be in the ball park. Plus the PCIe version of the 6600GT runs with memory at 1000 Mhz default vs 900 Mhz on the AGP version. I wouldn't be surprised if you were faster than that system by simply overclocking your memory to 1000 Mhz.

I also got a PCI-Express card and motherboard; on THG, they crossed the 8K line with a 3.2GHz P4 using a 6600GT PCI-e. I thought my processor is more powerful...

 

[boyRacer]

Originally posted by: Trente

Originally posted by: bR
Got a question for you... does your Acoustic Edge still work with most of the games you play? Philips abandoned driver support... but I loved that card.

Yes, it does. I was surprised to discover that it works flawlessly on my nForce4 platform. It's a shame that Philips abandoned driver support; It seems they actually dropped the number of voices the card can produce on the latest and last driver set (v1.89), as it fails the 60 voices test on 3DMark. Do you think it has anything to do with my low score?

Could it be that the Philips uses many CPU cycles that it affects my 3DMark score?

Maybe, I dunno... but I doubt it'll be anything drastic... i tested my onboard vs an Audigy and it couldn't do the 60 voice test either... plus i was losing around 6fps during the other tests but the score drop wasn't that big.