Is this the highest 3DMark2001SE score I will get?

[FluxCap]

Quick background:

I built my new system yesterday (listed below) and I have had very low scores (15kish) in 3DMark2001SE. Those are slightly lower than my Athlon XP 1700+ (overclocked to 1.83ghz). I got these low scores right after a fresh install of Windows XP and a fresh install of the Cat 3.9 drivers. I took some suggestions and checked the ram and updated the bios to the newest version and nothing improved. I installed older drivers (Cat 3.7s and Omega 3.7s) and the scores were still low. I messed around with a few more setting in bios and then decided to install the Omega 3.9s before I left for work. My 3DMark score went up to 16879. I was happy it improved but from what I have read this is still too low, right or not? Thanks.

 

[CraigRT]

it's less cause you are now using a pee four.. duh

seriously though, i don't know.. i found my stock XP2500+ scored higher than my 2.4B @ 2.9GHz. so maybe my above statement is correct

 

[FluxCap]

I hope that isn't true though. I was sure I would break 18 or close to it from what I read.

If anyone has a 3.0 and a similar system please post your numbers

 

[Richdog]

Originally posted by: FluxCap
Quick background:

I built my new system yesterday (listed below) and I have had very low scores (15kish) in 3DMark2001SE. Those are slightly lower than my Athlon XP 1700+ (overclocked to 1.83ghz). I got these low scores right after a fresh install of Windows XP and a fresh install of the Cat 3.9 drivers. I took some suggestions and checked the ram and updated the bios to the newest version and nothing improved. I installed older drivers (Cat 3.7s and Omega 3.7s) and the scores were still low. I messed around with a few more setting in bios and then decided to install the Omega 3.9s before I left for work. My 3DMark score went up to 16879. I was happy it improved but from what I have read this is still too low, right or not? Thanks.

Ouch, you should be getting higher than that most certainly. My rig nets me 17'800 in '01 and 6000 in '03. I would'nt have thought your CPU with all it's bandwidth would be that much lower than mine. I use the Omega 3.9's on full quality settings too.:beer:

 

[FluxCap]

I appreciate the response RichDog because that is the stuff I need to hear. I need to know that something is wrong so I won't just settle for sub-par performance

 

[Richdog]

Hmm now I think about it, while 15k was undoubtedly too low, 16.9k could be about right if you have quality settings on and your CPU isn't overclocked. Maybe the settings just need a little tweak to get optimal performance.

 

[FluxCap]

So, put the settings to performance in 3DMark2001SE and leave everything else the same and rerun it?

 

[modedepe]

Overclock your cpu/vid card.

 

[FluxCap]

Overclock your cpu/vid card.

Just because my numbers are low at stock I should? Don't think I would do that. Overclocking can come later after everything else looks normal.

 

[modedepe]

Well since you haven't overclocked yet your scores don't look that low. Just make sure you set everything to performance, no aa, af, or vsync.

 

[FluxCap]

I think that is my next step verifying that nothing is set incorrectly in display properties...

 

[Tom93R1]

Do you have hyperthreading enabled? You should get a better score with it turned off.

 

[FluxCap]

I have HT enabled, yes. I thought it really didn't make that much of a difference?

I finally adjusted a few of the display property settings and got a 16977. It is getting better but something like ram must be holding me back.

 

[Tom93R1]

I have noticed from running SETI@home and watching WU completion times that with HT each virtual CPU is about 2/3 the speed as just running it as a single CPU. Great for running SETI@home because I can do an extra 33% more work units per day, but bad for running apps like games that really stress the system. I havent tried 3DM2k3 yet, but I would imagine improving your processor by 33% will help things out a little.

 

[FluxCap]

Just ran 3 tests of 3DMark2001SE with HT disabled and it actually hurts the scores by an average of 55 points. Going to keep it on at all times now.

 

[Evdawg]

GAAAHHH! you people think 3dmark is allll that matters. Go download the UT2K3 demo, and benchmark. Its very simple and easy. Write down your scores.. and go find benchmarks somewhere on the internet. OR! just post them here and see what everyone else gets. I'm sure you will be beating more than 75% of the people here. It's not what you can do in 3dmark.. its what you can do in games. It's actually quite simple

 

[Camofrog]

Originally posted by: FluxCap
I have HT enabled, yes. I thought it really didn't make that much of a difference?

I finally adjusted a few of the display property settings and got a 16977. It is getting better but something like ram must be holding me back.

I have the same Mobo what are your RAM timings set at auto by SPD? you should check that and i found that if you have good memory this mobo sets them really conservatively when set to auto, so check what you ram is rated at and set it to that (since you is not O/Cing). May also want to go into BIOS and make sure you have "memory excel mode" enabled also maybe change it to "turbo" instead of standard or auto your not O/Cing anything by doing those things jus a little fine tuning i've figured out.

 

[DAPUNISHER]

Originally posted by: Evdawg
GAAAHHH! you people think 3dmark is allll that matters. Go download the UT2K3 demo, and benchmark. Its very simple and easy. Write down your scores.. and go find benchmarks somewhere on the internet. OR! just post them here and see what everyone else gets. I'm sure you will be beating more than 75% of the people here. It's not what you can do in 3dmark.. its what you can do in games. It's actually quite simple

Good post, a little melo-dramatic but good I agree too, but I'd suggest removing the 3Dmock scores from your sig so that the post doesn't come off as seeming hypocritical :light:

 

[FluxCap]

have the same Mobo what are your RAM timings set at auto by SPD? you should check that and i found that if you have good memory this mobo sets them really conservatively when set to auto, so check what you ram is rated at and set it to that (since you is not O/Cing). May also want to go into BIOS and make sure you have "memory excel mode" enabled also maybe change it to "turbo" instead of standard or auto your not O/Cing anything by doing those things jus a little fine tuning i've figured out.

My ram timings were like 4 4 8 8 at default which was terrible. So I changed them to 2 2 6 4 like somone suggested and gained about 1200 in 3DMark2001SE. I am consistently about 17900 now but people keep telling me I should be over 18k easy. I give up

 

[hemansd]

Originally posted by: Tom93R1
I have noticed from running SETI@home and watching WU completion times that with HT each virtual CPU is about 2/3 the speed as just running it as a single CPU. Great for running SETI@home because I can do an extra 33% more work units per day, but bad for running apps like games that really stress the system. I havent tried 3DM2k3 yet, but I would imagine improving your processor by 33% will help things out a little.

y would enabling HT cause gaming performance to go down while while improving other app's performance??

 

[Evdawg]

Good post, a little melo-dramatic but good I agree too, but I'd suggest removing the 3Dmock scores from your sig so that the post doesn't come off as seeming hypocritical

I didn't say they were pointless... i just said they werent everything. I think theyre cool and all, but theyre not everything. Thats all. My computer hasn't lagged yet in any game with every detail possible turned on, but i get lower marks than other people. I don't know where i was going with that... agh up till 5 last night seeing lord of the rings

 

[hemansd]

My ram timings were like 4 4 8 8 at default which was terrible. So I changed them to 2 2 6 4 like somone suggested and gained about 1200 in 3DMark2001SE. I am consistently about 17900 now but people keep telling me I should be over 18k easy. I give up

is there a program that allows me to check/change my ram timings?? i have a computer from dell so their bios doesn't allow me to do much

 

[modedepe]

Originally posted by: hemansd

My ram timings were like 4 4 8 8 at default which was terrible. So I changed them to 2 2 6 4 like somone suggested and gained about 1200 in 3DMark2001SE. I am consistently about 17900 now but people keep telling me I should be over 18k easy. I give up

is there a program that allows me to check/change my ram timings?? i have a computer from dell so their bios doesn't allow me to do much

You can't do anything with ram timings on a dell. You change them in bios.

 

[jiffylube1024]

Are you running dual channel DDR memory or is that single channel? DC should give you about 5-10% more performance on the P4.

 

[Tom93R1]

Originally posted by: hemansd

y would enabling HT cause gaming performance to go down while while improving other app's performance??

Because HT makes your OS think you have 2 CPU's. It never will allow a single application to fully utilize 100% of the CPU. With SETI I can run 2 work units (1 per virtual CPU) at a time and they both complete in 4 hours, or I can run one and it will complete in 4 hours, both these scenarios are HT enabled. If I disable HT the WU completes in 3 hours. HT doesnt improve any one app's performance, it allows your CPU to better run multiple programs at the same time.

The only reason I could think of that the 3DM2k3 scores would drop when HT was disabled is if 3DM2k3 is written to utilize 2 processors.