3DMark06 = 12750 --> Decent?

I just put together a build for my brother. He's running a micro ATX with poor flow on the CPU side, so we were limited with regard to OCing. The specs are as follows:

Gigabyte G33-based micro ATX
Q6600 OCed to 2.7GHz
EVGA Superclocked 8800GT (650MHz, I believe)

3DMark06 = 12750. I know this isn't a mind-blowing score, but is it pretty much in line with the above specs?

It is mind-blowing. It's among the best 7% PCs measured, maybe even better.

I'm a few points under 12000, with stock Q6600 (2.4 GHz) and overclocked 8800GTX (at 621 MHz). Your score is in line with the specs IMO.

Wow, that's gotta be in the top 1%

I wonder if a RAID 0 would increase your score to 20K or more.

29325 top score (I guess it's the world record?)

http://service.futuremark.com/compare?3dm06=3970826

I got exactly the same score with the system in my signature, my GPU clock is also the same 651MHz. Your score looks above average considering your quad is clocked under 3GHz.




By systemlord at 2007-12-10

Ive seen a 8GHZ celeron..... i know i know its a celeron, but still.

I to like to compare will people that I can talk with to get more information like you just got by all of us. The charts don't tell you that your 3DMark score has a lot to do with you processor or that even a wimpy 2900XT can beat any 8800 series card even if the 2900XT can't compete with the 8800 series cards in actual gaming. You'll have to get used to MrsBytch she is our...

My system scored a 12802 stock. Looks like yours is right on the money.

Q6700
BFG 8800 GT OC 625/900
2GB DDR2
Asus P5E

I've overclocked the CPU to 3GHz and video card and pulled out a 14470. I'm going to try to break 15k soon.

My Vista index is 5.7, I haven't seen anything higher than that before. It's kind of messed up that the 2900XT is scoring so high in 3Dmark, if only that power could be utilized in actual games.

Q6600 @ 3.0Ghz (should be able to take it up to 3.6 in the near future)
Patriot Extreme PC-9200 4x1024
Dual 8800GTS-512
16835 3DMark06 under XP x64

I like to use the 3DMark06 to help check if I have my rig running the best it can at whatever given OC I have. Look at your cpu and SM scores. 3DMark06 is heavily GPU dependent so obviously the better the GPU, assuming it is not help back by the cpu, the better the global score. My Q6600 at 3.46ghz gets a little over 5300 for a cpu score. The 8800GTS 640 SM scores are 5100 -5200. Globally I scored 13,044. Single GTXs and G92s score in the high 5000s to very low 6000s. Changing your drivers, cleaning out the registry, and shutting down nonessential processes can improve scores by a several hundered points or more if you are lower than similiar spec'd/OC'd systems. ie.My old FX60@3ghz/8800GTS 640initially got 8800+ after doing the above my score jumped to 9883. You can compare your scores against score scores of other CPUs and GPUs using TH guides. That should give you a idea as to how well your CPU stacks up to available chipset/GPUs, how much potential you have and maybe whether an upgrade is warranted. ie. At 5300+ my Q6600@3.46 is working much faster than a stock QX9650 at 3.0 that scores 4357.

You'd get a lot more if you could overclock the CPU slightly more. I get about ~13k with my E4600 at 3.5Ghz.

To answer your question. That is a GREAT score.

Well a stock Q6600 should be around 3600-3700...stock Q6800 @2.9+ghz is 4100ish so at 2.7ghz you are doing pretty good. The Yorkies run much cooler so their advantage is in their ability to reach much higher OCs. For that matter in games quad at the same clock speed are actually faster than the C2D's but the latter OCs much higher... so ultimately are faster. Obviously, benchmarks don't translate directly to real world use mainly because programs are not optimized to take advantage of the potential. It simply illustrates what the future may hold as it is begun to be utilized.

Very nice. Correct me if I'm wrong, but that Q6600 looks like it's running at 3.3 GHz (ie: far from stock).

You know, I really like ATI cards. I've been an nVidia guy for years, but only because they've outperformed ATI in raw numbers. If ATI can get a card out there that can perform within 5% of nVidia, with filters enabled, I'll be right there.

On the newest games can't go max settings and be playable, but one step down usually works great. The 320 MB of RAM on my 8800GTS probably has something to do with that.

Hoping for a graphics upgrade soon, seriously thinking about the 3870x2, however, dual GPUs always scare me. I still hear people complaining about the Geforce 7950gx2 and lack of driver support. Hoping the 9800GTX isn't as crappy as people are now forecasting.

anyone realize the 29k system uses two 2900xts?

This is prior to the OC in my sig.

Vista 32bit 11129 @2.4
XP 32bit 11757 @2.4

Still to test @3.0Ghz

Well, I just built up my system today: Q6600, 4 gig G.Skill, and an 8800GTS 512. Unfortunatly, my CPU is stuck at 2.8 GHz. If I try to OC beyond that, windows gives me a boot error. I've tried a number of things. It might be the RAM, I'm not sure.

Nevertheless, I was able to pull a 3dMark of 12865, which was over 100 marks better than my brother's PC, which has 2 gig, an 8800 GT Superclocked, and a 2.7GHz Q6600.

Why is that systemlord?? I have E6750@3.4 4 gigs DDR2800 and a 8800GT with a bench of 12640 on 3-D 06

http://service.futuremark.com/compare?3dm06=4599240

14470 on a Q6700 @ 3GHz and overclocked 8800 GT. Seems like your score should be higher.

Hey, well I don't know what to say, Everything is running stock except the CPU. I have had no luck at all usng overdrive to oc my cards. Not stable at all. When I had one in it was fine but once I went crossfire nothing but problems.

systemlord....many, many thank yous. I was about to pull the mobo and send it back. As a last ditch effort, I pulled one stick out and ~vollah~ I'm magically able to overclock past the 2.8 GHz wall. I'm currently at 3.6 GHz, 1.43125v (BIOS), 1.408v (idle; CPU-Z), and 1.376 (full load; CPU-Z). Do these voltages sound okay? Hottest core is at 66*C under full load (Prime95 small-FFTs).

The system is rock stable so far, with zero errors after over 5.5 hours of Prime95 small-FFTs. I plan on running small-FFTs for 12-16 hours total. I also want to test the remaining RAM, given that my original problem was a RAM problem. Which of the Prime95 tests is good for RAM? I've heard in-place large FFTs, and I've heard custom. Not sure which to go with.

I said "sometimes" overclocking with 4GB of Ram is tough, Some mobo's just don't like 4 sticks of Ram with a very high OC. Some of the best OC's in the world are done with only one stick of Ram. Example check out the post under this one.

You guys are going to laugh....

So I'm about to take a picture of the remaining stick of RAM for ebay, when I notice a little sticker on there "test voltage, 2.0-2.1" So I go onto new egg to check the specs, and I realize I was looking at the wrong bloody RAM! (listed as 1.8-1.9). So I just pop the other stick back in, up the voltage to 2.06, and it boots right up, at 3.6 GHz no less (doh!).

Temps in the house are a pretty constant 18-21*C all year round. I ran small FFTs for 17 hours, no errors. Figured out the RAM issue, added the extra stick, ran large FFTs for two hours and got a reboot. Pushed both the RAM and the CPU up one notch on the voltage. Large FFTs have now been running for 6 hours, no errors. I'm going to run large FFTs for another 2 hours, then run blend overnight. Assuming blend shows no errors in the morning, I'll run small FFTs again for another 8ish hours. If it passes that, I'll assume I'm stable.