XP 1800+ TB and 3DMark2001 SE scores >14000?

[screw3d]

Is there anyone out there who can score something higher than 14000 in 3DMark2001 with a thoroughbred 1800?

If yes, please post o/c stats, temps,bios rev. and system specs

The highest I've got so far is just about 13900.. but being a wuss i dare not push my system higher But I would like to know how high can I ramp it up.

The last time I tried a 180 FSB the CMOS crapped out n i hav to reflash it!

Thanks !

 

[Bovinicus]

You shouldn't compete for the highest score in a game you can't play. It means nothing. Get a higher score in Q3 with 4xFSAA and 8xAnisotropic filtering enabled and you're getting somewhere.

 

[BD231]

Bitch Please , lol . 1800+ T-Bred here, if you want more info check the rig in the sig.

 

[Darien]

Originally posted by: Bovinicus
You shouldn't compete for the highest score in a game you can't play. It means nothing. Get a higher score in Q3 with 4xFSAA and 8xAnisotropic filtering enabled and you're getting somewhere.

 

[BentValve]

There is nothing wrong with PC hotrodding, let the man be.

 

[HalfCrazy]

Here is my system specs and plus a url to a screenshot of my 3dmark2001se score.

3dmark2001se Score

AMD 1700+ Processor
Epox 8KHA+ (Rev 2.0)
512MB of Kingston PC2100 DDR
VisionTek Geforce4 TI-4200 128MB
Western Digital - WD800BB 80.0 GB @ 7200 RPMS
Western Digital - WD200BB 20.0 GB @ 7200 RPMS
17" Samtron 76V
Toshiba SD-M1212 - 6x DVD
TDK CDRW4800B - 48x24x48x
SoundBlaster Audigy
Windows XP Pro

 

[Ilmater]

Originally posted by: Bovinicus
You shouldn't compete for the highest score in a game you can't play. It means nothing. Get a higher score in Q3 with 4xFSAA and 8xAnisotropic filtering enabled and you're getting somewhere.

Are you kidding me? Why in the world are you concerned about getting 3000fps in Q3? Does that really allow you to play better? No. Plus, when was the last time you even played Q3? I can't remember the last time I did. You are in the same crowd that complains about using synthetic benchmarks in system reviews. Why? I don't use the REAL programs either! I don't care if it only takes you 70 seconds to WinACE a 40GB file because all I play is Age of Mythology.

No one benchmark is more or less valuable than another, as long as you get a general idea of how things are running. And with reviews, it's more important to vary the benchmarks you do than to do certain ones. Athons have better FPUs, so anything that involves serious number crunching will get done faster on them than on P4s. But, on the other hand, P4s always wipe up on the Lightwave benchmarks. Who cares?! Someone who uses Lightwave or crunches numbers (note that the next Cray supercomputer is using ~10,000 Opterons), that's who! So if he plays 3D games, then using 3DMark to bench his 3D performance is not only his perrogative, but a good idea.

btw, nice score!

Edit: I also want to add that last night, I benched my computer, having just recently done a fresh install. It got 1000 3DMarks LOWER than it did before I did the fresh install. Are you telling me that my computer isn't actually performing slower than it was before? Of course it is. So isn't that program, then, a valid interpretation of true performance?

 

[screw3d]

You shouldn't compete for the highest score in a game you can't play. It means nothing. Get a higher score in Q3 with 4xFSAA and 8xAnisotropic filtering enabled and you're getting somewhere.

I m not competing against anything. The benches are meant to determine the maximum possible o/c without compromising stability. If the point is just to get the job done, why do people prefer driving beamers n mercs instead of just a lousy chevy from the 60s? U'll still get from point A to point B rite?


BTW BD231, what's the FSB on ur rig? that's a shweet score

 

[BD231]

200mhz FSB . 200mhz FSB is a thing of the past now though, seems like most everyone is getting there. You should be able to get to 200mhz fsb no problem, we have identical rigs!!

Set CPU interface to Aggressive

Mem timings should look like this:
7
--------
3 Note these 2 settings NEED to be at 3 or above, any lower and my 8RDA has sound and crashing
3 issues. Overclock is limited as well.
--------
2

Lock your AGP port at 66mhz

My voltage is set to 1.675v, which is fine with the cooling I have. I tried for the default 1.65v that these processors run at speeds in excess of 1800mhz I think it is, but it needs 1.675 because I think the 8RDA undervolts AND overvolts a bit. When I'm set at default v-core it over volts, and anything past 1.60v it undervolts . Anyway, you should be able to get that chip to at least 1900 with those mem timings and voltage settings. As you can see the mem timings I have do not effect performance all that much.

 

[screw3d]

Hmm.which would u consider more important : Faster FSB or better memory timings?
Right now I'm on (only) 166 x 10.5, RAM at 4-2-2-2...

At this memory timing I dont think I can go beyond 175.

Is there a noticable difference with different memory timings?

 

[KickItTwice]

The most insignificant benchmark you could ever possibly use would be Quake 3. When Quake 3 first came out it was a very good way to determine how future games would play on your new video card, because a lot of games coming out would be using the Quake 3 engine. That is now the past. Any video card you buy today that is worth anything will own Quake 3. The 200 - 300 FPS your getting is all about CPU limitation. If you wanted to single out a game to use for useful benchmarking, then it would be the UT2003 benchmark using the outdoor level ( Antalus ) to test your video card. 3DMark is the absolute best benchmark because it uses a variety of games, not just one game, as well as some other tests to give you a very well rounded score that represents what you can expect from your system graphics wise. When you look at video card reviews where they run tests with several games, you can see that certain games favor certain hardware. The 3Dmark benchmark is more of an average of using more than just one game to give you more of an average score, though 3dMark does include tests using features that are not out in games yet. I find this to be a good thing because when I buy a video card, it's nice to know that it will play games coming out in the future using advanced features like pixel shaders.

Crap, It's time to go to work. Later...

 

[BD231]

Originally posted by: marvie


Hmm.which would u consider more important : Faster FSB or better memory timings?
Right now I'm on (only) 166 x 10.5, RAM at 4-2-2-2...

At this memory timing I dont think I can go beyond 175.

Is there a noticable difference with different memory timings?

Nope, with those mem timings you run the chance of killing your HD :Q. Read write errors can occur which can completely screw a hard drive, or give it bad sectors(it all has to do with that first timing, it should never be put under 6. Look at my edit in the post above, that should get you to 200mhz fsb and possibly further. As long as you stay at cas 2 the rest of the timings don't make much of any difference I've found.

 

[screw3d]

How does too-low timings kill the HDD?

Anyone has experience this b4?

 

[screw3d]

200mhz FSB . 200mhz FSB is a thing of the past now though, seems like most everyone is getting there. You should be able to get to 200mhz fsb no problem, we have identical rigs!!

I am having problems :frown:

I just can't run at 200..so I'm still stuck at 175... at 1.84GHz

I've tried to do whatever that you've done but

Wonder what's the problem? I suspect the RAM but it's a corsair xms so it's not too likely..

 

[BD231]

Low timings can cause write and read errors in your memory which can lead to issues with your HD, I'm not saying it's a sure thing and that it'll happen to you, but it dose happen. There are people on these boards who can attest to killing HD's by overclocking. Have you been droping your multiplyer and making sure your AGP port is set at 66mhz?

 

[screw3d]

Yeah I am pretty sure that I've got my AGP clock and the memory timings right, as u say, 66Mhz AGP n 7-3-3-2

But when i run at 200MHz x 9.5 I can't load Windows XP it'll just reboot.

Anything else?


Oh and another thing.. does your network's address (the ipconfig /all thing, the format is like xx-xx-xx-xx-xx-xx ) change everytime you fail an overclock? mine does..

 

[BD231]

I don't mess with the network address thing, I just have it on auto. what voltage are you running at?

 

[screw3d]

It's not a network setting that you can change. I think a NICs' physical address (not IP address or anything like xxx.xxx.xxx.xxx) is permanent. Maybe cos the NIC is built-in the chipset, it's sorta changable through software? I need some explanations on this

I tried 1.675 as u said.. but it still won't work. Right now I've put it back to the default 1.6V

 

[BD231]

I don't use onboard lan, so I can't really explain that for you, I've never taken the time to check if mine changes. You should be able to get to 200mhz fsb considering your memory is rated at 200mhz(400ddr). Maybe all the on-board stuff is causing the problem, in any case, try lower settings (2.5 cas) and see if that works, if it dose then you know it's the memory holding you back. You can also raise the mem voltage, I wouldn't go past 2.73 though. In any case, just keep trying different settings to figure out what's holding you back.

edit: DOH!!, set CPU to optimal . Epox has a VERY vauge discription on that setting and I think I got it mixed up. It says Optimal: For most Stable CPU/FSB settings, Aggressive: For overclocked CPU/FSB, horrible way to describe it if you ask me, hat last one got me thinking that aggressive was FOR overclocking specifically. Good luck.