der8auer

Nvidia GTX Titan reference PCB Pics of the card: Mods: Also, please be aware that the reference VRM design by Nvidia is really weak once you start overclocking. As Rbuass mentioned earlier, his card died at a mere 1.46V Vgpu! If you want to go extreme, you should look into your own zombie vrm. These are the modifications that are on the pictures: GPU (NVVDD) Voltage Measurement GPU (NVVDD) OverCurrent Override GPU (NVVDD) OverVoltage GPU (NVVDD) Coldslow / shutdown GPU Core Clock Range Shift Up Memory (FBVDD/Q) Voltage Measurement Memory (FBVDD/Q) OverVoltage PLL Voltage Measurement Front and back pics (click for high-res version):

No, for G80 check here http://forum.hwbot.org/showthread.php?t=93508

ASUS 8400 GS silent custom PCB Pics of the card: Mods:

Nvidia 8500 GT reference PCB Pictures of the card (might also appear with a green or blue PCB) Pencil Mods: vMEM & vGPU Mods:

Gigabyte 8500 GT custom PCB Pics of the card: vMods:

vGPU vMEM

Nvidia 8800 GTX/Ultra reference PCB G80 Pics of the card: vMods: For the 8800GTX the voltage for the G80 GPU and GDDR3 memory can be read out at the following places: Just like we showed on previous page you might want to solder a couple of wires to those readout spots so that voltage reading is easier once your card is being installed in your housing. The default voltage for the GPU is 1,3V; the default memory voltage is around 1,95V. GPU Voltage Modding You need a soldering mod for this one, therefore one must connect a 500 Ohm variable resistor between the marked spot and any ground. To increase voltage you must lower the VR's resistance, so always make sure that your VR is set to highest value before you turn on your pc. MEM Voltage Modding For the MEM voltage mod we must search for the small AT PWM IC. The marked resistor is the one you need to shade with your B2 pencil. Strike it a few times and measure how the resistance changed. You can also try the soldering mod, therefore one must connect a 20 kOhm variable resistor between the marked spot and any ground. And again you must lower the VR's resistance to increase voltage, so always make sure that your VR is set to highest value before you turn on your pc.

Nvidia 8800 GTS reference PCB 320/640 MB G80 Pics of the card: Mods: You might find it to risky to read voltages at those places, if you slip you might indeed damage your brand new VGA card. Therefore we soldered 2 wires on our card which are much easier to touch during our overclocking article. We used a small tip of glue so that the wires stay in place, this might avoid accidentally contacts between each wire. GPU Voltage Modding For the GPU core voltage mod we must search for the Primarion PX3540 PWM IC. The resistor marked in blue is the one you need to shade with your B2 pencil. Strike it a few times and measure how the resistance changed, you can check resistance by holding the red wire of your multimeter to the red point and the black wire of your multimeter to the pink point. The stock resistance measures around 35 Ohm, this is very low and this will make it very hard to pencil mod the GPU core voltage, we recommend doing the soldered modification. For the solder mod we used a 1kOhm adjustable resistor (often called VR) and soldered it between the red and pink point. This changed our stock voltage from 1.3V to 1.33V. To further increase voltage we must lower the resistance of the variable resistor, before you power on your system always make sure that your VR is set to highest value. For 1,4V we adjusted the stock resistance of 35 Ohm to 29 Ohm. MEM Voltage Modding For the MEM voltage mod we must search for the ISL6549PWM IC. The resistor marked in blue is the one you need to shade with your B2 pencil. Strike it a few times and measure how the resistance changed, you can check resistance by holding the red wire of your multimeter to the red point and the black wire of your multimeter to the pink point. The stock resistance measures around 536 Ohm. For the solder mod we used a 20kOhm adjustable resistor and soldered it between the red and pink point. Doing this changed the voltage from 1.91V stock to 1.95V. To increase voltage we must lower the resistance of the variable resistor, before you power on your system always make sure that your VR is set to highest value. Here is how we modded our card: For the MEM mod we actually didn't use a 20k VR, we made the combination of putting a 10kOhm VR in series with a 10kOhm normal resistor. This will avoid over voltage (max 2V) and more currency on our voltage range. For both GPU and MEM voltage modifications we used the same 'ground' connection. If you watch on our mod-pictures we see that we always have to solder between red and pink points. In real life those red points are points connected to ground. You can use any ground point for doing those modifications, though we always made sure we had a ground point available on each picture for people who want to use those points. OCP Mod: I Have successfully made the OCP mod for the 8800GTS (G80). What you do is solder 1.2K SMD resistors across the 3 caps i have outlined in pink. The stock resistance across the caps was 1.29k and with the 1.2k SMD resistors on it is reduced to 0.61k. This reduces the OCP signal by 51%. I also realised during testing that it also reduces vdroop from ~0.08v to 0.02v under load. I successfully ran bench after bench @ 1.52v idle, 1.50v load. You will need to improve cooling though, i have a modded heatsink & fan and it still hit 93*C during repeated runs of 3dmark05 GT2 but it was totally stable. I was able to bench stable at 702MHz gpu, the previous max was 684, and theres more left in it. Source: http://www.xtremesystems.org/forums/showthread.php?171016-G80-8800GTS-OCP-VDROOP-MOD-Made-amp-Confirmed-Working

Looking for a Rampage III Extreme Black Edition. Let me know if you have one for sale or if you know a place where I can get one. edit: Now looking for a REX X48 LGA775

whahahaha really lost it looking at those pictures

Yea I remember I already forgot it until I started benchign - Wait there was something... We failed anyway yesterday.

No backups here. We already uploaded our best score 3 weeks ago. So who's left? AU and US? No matter what - AU won the CC 2013. Well deserved guys. Congratulations! edit: According to Massman Australia will be known Sandstralia from now on

Lawl guys still stacking up the bags?

Thanks a lot guys!

I'm just benching 4 way SLI with 780 and we have the bug that PhysX is always stuck for the first 2 seconds and then continues normally. The result is a very low score in physx with about 19k at 6 GHz where we should have 2-3k more. Does anybody know how to solve this? Funny is that at 5 GHz there is no such bug. But at 5,1 GHz it starts bugging.

Actually great Wanted to bench tomorrow morning. Now I can sleep longer.

8.1 is allowed but you need the latest sysinfo and you have to provide a valid futuremark link. 13.12 is fine, too. All 3DM11 versions are okay except 1.0.0

There are a tonn of different cases and I'm pretty sure you guys are sometimes also happy that we investigate each case individually. In the case above the result can be tricked without the correct system info on Windows 8. However the Cloud Gate results are the same on v1.0 and v1.1 using single GPU. http://hwbot.org/competition/country_cup_2013/stage/1158_superpi_32m_low_clock_challenge/ Stage limitations: Engineering samples and hardware released after November 1 are not allowed

People using v1.1 did not have any disadvantage so there is nothing to complain about. It would be different if 1.0 scores higher. I can't help if you did not manage to submit 3/3 results. But I really disrespect the behavior of reporting submissions just to go up in the ranking yourself. End of this discussion for me now.

It seems like you don't see a lot of points. If you're the rules-hero here you should know about this part: The submissions only have a different version which does not affect the final score so there is no reason at all to block them. Even if they are against the rules. Those competitions are for fun and there is absolutely no reason to disappoint so many people for no reason.

I checked those results now: http://hwbot.org/submission/2470180_bmbenx_3dmark___cloud_gate_geforce_gtx_570_29127_marks http://hwbot.org/submission/2470144_yudhiagust_3dmark___cloud_gate_geforce_gtx_670_32927_marks http://hwbot.org/submission/2470074_dinos22_3dmark___cloud_gate_geforce_gtx_570_34023_marks Please let me know if I missed a result.

Good news guys Submissions will be linked back to the competition.

I just really dislike the fact that you reported the submissions few hours after the contest was over and I guess you did not just realise this at exactly this moment? Real sportsmanship would be to let guys know that there could be a mistake. Especially if there are a lot of results and you apparently went through every single top result and checked the version and reported it. My GTX 570 result (first setup with 3770K and therefor lower score) was uploaded on 2nd of November and I already used v1.0 there. That's a timeframe of almost two months.... And now it turns out that the issue with v1.0 is multi GPU related and we did a single GPU benchmark. I don't even find words for that really. Christian Ney is head of Moderation and has the final decision. I texted him this morning and he said he'd look into it this evening, so it's just a matter of hours. I'm pretty confident tho that the results will be moved back to the competition as the 1.0 "issue" is just affecting multi GPU scores.