I.nfraR.ed

Are you sure vCore is that high, have you measured it with DMM?

You're almost safe, mate. I don't have time for LN2, too. And I don't even have SS, so will stick with water. Don't think I can pass the 50K mark with what I have. Plus all my good cards are insulated .

Not only the motherboard, but also ram and cpu.

I had two of these, but damaged the memory on the modded one in a LN2 bench session. It was running 1400MHz DDR3, but I accidentally shorted something and gave the memory 4V . This one is unmodded yet. Gainward uses 0.8ns Samsung chips on the GS . http://hwbot.org/submission/2312149_i.nfrar.ed_3dmark03_radeon_hd_4850_gddr3_69158_marks Yes, it's the stock bios with stock clocks on Gainward HD4850 GS.

No, you wrote at the time I was writing...it was addressed to the previous post(s) or in general for that matter... .

So you say if I only could replace a 3DMark03 scene runtime with another scene which includes a plain perfectly flat land and sky, it would be legit? And get 1000% boost? Or maybe omit the sky too, why not...

Does anyone know how to make it detect the real cpu frequency? Maximus 5 Formula + 3770K. PS: I think showing the detected cpu clock in the screenshot should be mandatory. We all know this is the culprit for the benchmark. I try to make it detect the real frequency, not the stock multi x bclk, but this is the place which is the biggest loophole of '99 and '00.

Load optimized defaults, set sata mode to ide and try again. I had this problem on newer M5G bioses and now on M5F. Still under investigation what exactly causes this error, but I'm able to install stripped SP2 this way. I have the ssd plugged in one of the red sata ports. You may also tweak the settings under Monitor tab and disable "Monitor Standby", they don't have effect on the error. Btw, I had this problem installing Windows 7 as well. The forementioned method worked.

Not impressed, I can do better... joking 32M master, that's it.

Great cpu and props for that vcore!

Tapakah, I don't have more suitable board . My Crosshair IV is coldbugged and I sold CH3 2 years ago. Tried wprime with this one, but is a no go. Beep, I'm not very satisfied with the result, but it's still under 10s, which is fine . With my usual eff, it should have been 9.83x - 9.84x range. 32M is not very good either. Plus one dream came true - 7GHz with Deneb, although it was pretty hard. Also thinking of buying some new chips while they are still in stock.

When I had 6.6GHz chip, a 6.9 one looked huge in my eyes. Now when I tested 4 chips which did almost exactly the same 6.9GHz +/- 10MHz...I want 7+, lol.

I've been extremely lucky lately. Maybe I need to buy more chips and then make some money

Going to try it in a while, need to insulate the board first. I have a score to improve .

What about 1.9+ V and 1 core, does it scale?

First two sound like 7+ GHz chips, Splave. Hopefully they will scale past 1.84V, then you have a winner!

No, I made a "fake" pifast. Simple c++ project for the sake of showing it running, otherwise a simple bat file would output whatever you want just for the screenshot. This benchmark is the most insecure and have a lot of loopholes. Strange I haven't thought about this before. What I did is illegal ofcourse. But I made it obvious. If I had bad intentions, I could do a "normal" result that won't be suspicious. I don't think everyone can do it, because the specific way pifast calculates the times, but it's not that hard.

Yes, I already suggested that long time ago, so I'm up with you - VGA slots, CPU sockets.

One with a large cache Remember Opteron 144@3.1GHz beating Venice at 3.4. Not sure about dual-core vs single core. Dual core should be better, but if it's too gpu-bound, I don't know. Finally some use of my AGP cards, but I'm not sure if I will have time for all this.

I found some great tweaks for pifast. They have similar effect of the super tweaks in PCMark05. Managed to beat the current record with cpu at only 5.1GHz. Before: After the tweaks: Video of the tweaked run: Don't know if my calculations are right, I just wanted to show it's relatively easy to fake a pifast score without photoshopping it. PS: This is not really related to the case here and the screens from John, but thought it would be worth showing.

I don't want to sound like a hater, but if it was me or some other ordinary overclocker (not in the pro league), I'd have been banned already. Those evidences on the screenshots would mean direct ban. Don't want to ban anyone without reason, don't get me wrong, but it shouldn't take that long to investigate the case. I understand that it could be bugged run or some glitches on the screen (regarding the cut label on the shortcut). It could be similar to when some parts of the baloon tooltip are sticking sometimes on the screen or the asus turboevo transparent area is making an invisible overlay on the screen. But it should be a user responsibility to check every screen he's uploading, especially when some of his submissions were questioned recently. However I can't explain myself the odd text displacement/errors in the cmd windows. It's clearly not a valid run, whether bugged or shopped.

why not I've gone up to 1.5V (didn't scale with more) on normal ambient water cooling and 1120MHz gpu.

I can eventually try with my GTX2, but they aren't matched pair anymore. One of the sticks died (dunno how and why) and now I have 2 different sticks. The matched pair ran well on the Maximus IV Gene and Sandy, but required a lot of tweaking.

This one - no. Only pifast and 1M at ~1300, but when I tried 32M it was crashing on the initial loop. On the other 3218B cpu I had no problem at 1300 on cold.

You guys suck at Photoshop - missing superpi, different font, no alignment, etc. CN, ban them all! Here's the correct one and all mu subs are shopped as well.