Originally Posted by Masked
Lots of words.
Look, I am not sure what you are wanting.
You want credit for being the first one to notice it? It's all yours.
You want credit for being convinced there is a physical problem? It's all yours.
You want me to believe there is this some nVidia conspiracy-esque issue? Well, I am not going to, but if you want to tell people I agree with you, go ahead.
We had a saying when I was in the military:
If you are not a part of the solution, you are a part of the problem.
As far as your high and mighty stance over me:
You give no conclusive evidence there was a physical problem.
You have given no evidence that nVidia knows about this.
You have given no evidence of any thing about anything essentially.
You also claim this is apart of the OCP. I went into the BIOS, and there is NOTHING relating to the OCP on this issue. Voltage is simply just fed wrong to the GPU it seems.
One thing I don't tolerate is BS from anyone. I don't care if you are the president of Truth.inc, I will call it how I see it.
This is the last I am talking about this though
All I have done:
Essentially, I discovered on my own there was a problem with CUDA based application. I traced it back to voltage on GPU2. I tested a hypothesis. Solution worked. That's it. Very simply. Problem solved so far.
I am still going strong for nearly 36 hours continuous GPU computing, with 0 errors. My last error was the 24th of April
You can see for yourself:
Lots CUDA units failed that day.
So far the stock clocks set to 0.925v is working perfectly. Rather than accepting that I may have come up with a solution, you suggest I am blind, refer to Khadafi, and other stupid remarks.
You claim all of my sensors are wrong, yada yada yada, and then I go and prove that your AIDA64 has had a bug with it, that didn't display the correct voltages/GPU usage for GTX 590. Then you claim thats the one you are using, you have been using this released beta for-ever, even though it was only released like two weeks ago. So even if all my sensors are wrong, the cards are working top notch on high-end physics based computation work units with 0 errors. But that doesn't matter because you don't care about BOINC, and you are an expert Alienware administrator, who builds elite and private CUDA applications, and has a private army of GTX 590s at your "disposal."
To be honest Masked, just about every post you bring up, I don't believe about 95% of the things you claim. I don't have to, and you could be telling the truth. So like I said, I am not talking about this anymore, its just going to drag on and on. I personally don't like drama, which is why I asked you to stop derailing my EVGA thread with all this over-zealousness.
I apologize for arguing this way, sometimes you just have to take a stand, and pull out some numbers for people who have lost their way. Just like I do when people say the GTX 590 doesn't overclock.
I wanted to add I saw this fantastic 6990 CFX vs. 590 SLI.
I was shocked that the GTX 590 pulled ahead.
I will pm iceblade.
Originally Posted by krazyatom
Hey guys do you guys recommend gtx 590 in air cooling? I have no plan to go water cool.
My 6990 is way too long and it won't fit in my case
I might grab gtx 590 because it's shorter and it will fit. I am tired of waiting for gtx 580 3gb.
Yes, its moderately well for air, however, you need a good airflow case I am not familiar with yours, but if you are able to get a steady flow of air in the lower region, and exhausted out the back/roof, there will be little to no heat issues.
Looks like TiN.... has begun. There go my benchmarks...
Edited by RagingCain - 4/26/11 at 9:20am
Each GPU powered with separate 10 phase VRM, capable to feed up to 2V.