Benchmarking Contest!

So are there any improvements with the scores? What was your highest?

 

Nope. I was able to get the 10.6 drivers installed. Had to jump through a bunch of hoops. Scores went down after that. I need to try new drivers yet again. Unfortunately I'm working today.

 

Since you wanted to see some score (I do have my system with me at work) I figured I would post some since I now can change my drivers. This is with no additional cooling and just adjusting the multipliers.

 

Thats impressive, what do you think might be causing my system to be unstable at 28 multi? I had an issue with the screws being loose which caused the CPU to overheat, and then i reapplied the thermal paste. it then started throttling again, and i open it up and the thermal paste was ruined, do you think i might have torchered it a bit too much? Also, are you increasing Vcore?

 

Yes I raised it to 150. It's maxed out as far as the CPU frequency is concerned.

Regarding your situation I don't think it was something you did to stop it from hitting 28. I know you said you put the system in a fridge then benchmark but what kind of cooling are you doing during the benchmark?

 

The last time it was in the fridge, 06 would crash on 28 multipliers unless i decreased the Bios overclocking to 3% which gave me 21.5K. Now im running it with a normal fan, i had laying around the house, Combined with two small ones that are exactly under the GPUs. BTW, when it was in the fridge i had the same fans.

 

Which drivers are you using again.... 10.3?

Also how do you know it's the multi that is crashing 3DMark06? Have you tried it without OC'ing the GPU?

 

10.3 is the one im using, I dont remember running it without Ocing the GPUs but it would usaully crash in the CPU test.

 

I would still verify it.

 

At stock GPU multipliers at 28 + Vcore at 150Mv. I got a BSOD whilt running it with this error dxgmms1.sys (if i saw it properly). I googled it and some other forum members said it was due to DirectX !?? It crashed in the first GPU test, so is this a driver issue?

 

More than likely. Try using the Dell OEM 5870 drivers on their website and see what happens.

 

5870? I found them, and im downloading them. The weird thing is, i used driver sweeper to remove the drivers, and it just will not install the ones on the disk.

 

lol, I have the opposite problem. Only the Dell OEM drivers work unless I'm willing to install each component separately and manually.

 

I ran the benchmark on 27 multipliers, any thing higher would give me a BSOD with the same error message. LINK:- http://img249.imageshack.us/img249/5227/runb.png
Mandrake, im just wondering how the BCLK, went up to 143.4, if i never set it to anything in the first place? Why does it downclock the multipliers to 25, if i set them to 27?
I know the score sucks, so i am rerunning it again, with a few tweaks to decrease temps from CC.

Edit: whilst running it again i noticed the BCLK went to 149 and the multipliers decreased to 24.
As far as i know the system does not go past 142 on the BCLK, Maybe thats why abo cant support the system LOL

I found this old screenshot i had, where the BCLK went to 153 http://img94.imageshack.us/img94/9391/tempqd.png

 

I don't think hwinfo is reading it correctly. Have you ever seen evga eleet pop up with those numbers?

 

Hwinfo was open during the benchmark and if i close Hwinfo and open it again it shows 133 again. evga usaully shows 132.9 So does that mean it keeps on changing when the processor is being stressed?

 

No I don't think the bclk frequency is dynamic. That would cause too much instability and fry some memory modules along the way. I think the issue is with hwinfo. The only thing that should be dynamic are the turbo mode multipliers.

 

Just to make sure, is this happening to your system as well?

 

I haven't checked. I only use hwinfo for temps these days. I can check later on today.

 

Do you think it might be possible to disable turbo mode in the bios and increase the multipliers through windows?

 

never tried it but I guess it wouldn't be too difficult to test.

 

I just checked mine and hwinfo is not giving me the wrong bclk frequency.

 

Does that mean there is something wrong with my system? I had a feeling it might have been the ram overclocking so high that it gets unstable, i guess that the proof but how can i make sure? It increased by increasing the multipliers only, that probably why it was only stable at 18 multi and 3% BCLK increase. Any thoughts would be greatly appreciated. BTW, you probably see how its impossible to get to rengsey's score

 

I only have problems because of system instability but not do to OC'ing. My M17x has to go back. Too many failing parts in the first week.

Regarding your situation you need to see if another program is reporting the wrong bclk freq. I think the problem is with hwinfo.

 

Failing parts? I hope you can get a new one soon BTW, shouldn't you hit 22K with 29 multipliers ?? rengsey got his high score with 28 and 5%, looks like we will never know what he did to his system

Do you think removing one stick of RAM would decrease the instability?

 

No, I don't think it would have any affect.

 

Theoretically, what if we took an Alienware m11x, hooked it up to a nuclear reactor to power it, then coated it with arctic silver compound, and threw it in a vat of liquid Helium stationed at the South Pole? We could dismantle the outer shell and disconnect the LCD screen "just in case." Project Chernobyl as it were lol

 

We'd find it in a million years still intact.

 

Right next to USS Voyager, in the ice.

 

Is Wattos the only one trying??

 

Not many people have the 920xm. If I had it, of course I'd be trying; who doesn't like winning . But without that sick cpu, theres no chance

 

I would try but I already have one of those xps bags.

 

I am going to try after installing xp. You can always have two of those bags claxdog If xp doesn't increase the score alot more, then its just hopeless to get to get to rengsey's score as i haven't seen many hit 21K other than you mandrake.

 

22K is well within range. You need to find the right power usage for the system for the ultimate score. If you are over-volting your cards you might want to back down on the CPU a little and vice-versa. Need to keep testing to find out.

 

Thats what I do every day so are you getting another m17xr2

 

I cant get mine the bench at 28x get bsod every time due to temp.tech is coming to replace motherboard tomorrow due to high pitched noise(not gpu )so i will apply some paste while it is open.Ill see what happens with temps after that.

 

I will definitely be trying as soon as i get my clevo w870cu, i have a 920xm to drop into it and have a 180w power adapter as well.

 

so far i only got to 16.5k still some ways to go. i don't need the bag can you just gave me cash if i win,lol.

 

neither can I,try using less BCLK increase, i used 3% + 28 multi

 

don't know yet

Sure I will give $50 instead of the bag if you wish. One thing, all cash transaction have a $100 processing fee. Good luck and I hope you win!

 

Bump this up. Hopefully there will be some 5870XF entries soon otherwise I'm ending this contest.

 

Last call??

 

It's calling crippled grandpas on a running contest. Good luck ^^

 

Yeah, its tough because not everyone can afford the best processor. Heck, I only have the i7-720qm in my m15x and that was a huge stretch for me. Cant even take the laptop anywhere cause I dont have a bag lol.

It's an interesting contest for sure, but the barriers to entry are a bit steep for most people.

 

This contest is for 2% of the users on this forum lol

 

Well i tried, but without some proper options in the bios, i was not able to beat your score. I managed to get ~16,000 though, which i thought was respectable considering i cant OC the proc like you can on your g73.

 

The M15x cannot either but owner hit 17K.