Hi Everyone,Please do not run Kombuster, Furmark or OCCT, or any other benchmark program that gains control over the voltage setting in your equipment. According to both Nvidia and AMD these programs, as well as others that they refer to as "power viruses," will destroy your video card. As a matter of fact nearly all the benchmark testers have ceased using them altogether. In addition, the 580 GTX now has a built in alteration that detects these programs. and throttles back the power.Here is a link that explains it in a bit more detail:No Power Viruses AllowedKind regards,

I wonder if anyone here fried their card using this thing... A look at the OCCT website it flashes the Nvidia logo around like it's official.

The issue is not a new one and actually stems back to early 2009 when the developer for OCCT exposed a flaw in 4000series ATI cards see here: http://www.xtremesystems.org/forums/showthread.php?t=224958I think it was simply a prudent step that NVIDIA recognize a characteristic in their cards and take steps to protect them in their programing.The issue is not any different than throttling the CPU to protect it from damage, so now the GPU guys are on the same boat.OCCT and Furmark are still good and valuable test tools when used by knowledgeable people. GPU-Z even has built in a feature to disable GPU throttling for those that wish to continue to test their GPU to the maximum, beyond that of normal games.Overclocking can be harmful to anyone’s hardware and if you do not know what you are doing you should not be doing it. For those that want to continue to play you are encouraged to play responsibly and not burn your houses down.

The point is no benchmark should even attempt to access your voltage settings... definitely a virus. No application should increase voltages period.

Yeah, cranking up the voltage to test the card is like having an HVAC technician light your curtains on fire to test how well your furnace heats the house...Bob ScottColonel, USAF (ret)ATP IMEL Gulfstream II-III-IV-VColorado Springs, CO

I'm not sure where the idea developed that programs such as Furmark somehow have control over the voltage of your graphics card, but it is not correct. What they actually do is generate a load which causes the graphics card to draw more current than is safe, hence Nvidia's corrective action with the 500 series. A much-needed and wise move, IMHO.

I'm not sure where the idea developed that programs such as Furmark somehow have control over the voltage of your graphics card, but it is not correct. What they actually do is generate a load which causes the graphics card to draw more current than is safe, hence Nvidia's corrective action with the 500 series. A much-needed and wise move, IMHO.

But a very odd move, IMHO. If overcurrent is unsafe, why would they program the driver to only do current limiting when it detects that one of a few utility programs is running? Are we to take from that that overcurrent when running Crysis on two 30-inch monitors at hi res with 8x FSAA is OK, but running OCCT isn't?I stopped using the OCCT GPU stress test quite a while ago, as the load was obviously far exceeding that of any program I could throw at it...with FS or any recent 3D game with all sorts of heavy AA I could get the GPU to hit 70 deg C, but OCCT's GPU test would slam it right up to where I shut it down at 80 deg C. I don't know what exactly is going on behind the scenes...if the driver/firmware is increasing the voltage to control an overcurrent condition (my guess as to what is really happening), or if, as some would suggest, the test program was manipulating the voltage itself. But clearly the GPU test was not representative of any GPU load that I could anticipate putting on the card.That said, I still use the OCCT CPU stress tests as a staple in my testing suite.RegardsBob ScottColonel, USAF (ret)ATP IMEL Gulfstream II-III-IV-VColorado Springs, CO

But a very odd move, IMHO. If overcurrent is unsafe, why would they program the driver to only do current limiting when it detects that one of a few utility programs is running?

Graphics cards are such complex and sensitive devices, it only makes sense to have a multi-tier power-limiting system in place to prevent damage to the card. In the case of OCCT, Furmark and the like, these applications are known to generate workloads which could be dangerous to the card. Since there is no way for the card (hardware) to know what applications are generating its workload, this is where software comes in and pre-emptively limits power draw. For potentially dangerous yet unknown workloads there is the hardware over-current protection.

Are we to take from that that overcurrent when running Crysis on two 30-inch monitors at hi res with 8x FSAA is OK, but running OCCT isn't?

I wouldn't be concerned about it if running current-generation hardware from either vendor. Even with older hardware that doesn't have hardware protection mechanisms, if you're running a relatively recent driver you've still got the application (exe) detection that will prevent these specific programs from causing damage to your card.

I stopped using the OCCT GPU stress test quite a while ago, as the load was obviously far exceeding that of any program I could throw at it...with FS or any recent 3D game with all sorts of heavy AA I could get the GPU to hit 70 deg C, but OCCT's GPU test would slam it right up to where I shut it down at 80 deg C. I don't know what exactly is going on behind the scenes...if the driver/firmware is increasing the voltage to control an overcurrent condition (my guess as to what is really happening), or if, as some would suggest, the test program was manipulating the voltage itself. But clearly the GPU test was not representative of any GPU load that I could anticipate putting on the card.That said, I still use the OCCT CPU stress tests as a staple in my testing suite.

I like OCCT's CPU test also, I use it to verify stability of all my CPU overclocks. Particularly the Linpack x64 test. As for OCCT's GPU test generating a non-realistic workload, I agree for the vast majority of applications. However, running GPGPU applications such as Stanford's Folding @ Home, one can achieve very high levels of power draw that approach those of OCCT.

But a very odd move, IMHO. If overcurrent is unsafe, why would they program the driver to only do current limiting when it detects that one of a few utility programs is running? Are we to take from that that overcurrent when running Crysis on two 30-inch monitors at hi res with 8x FSAA is OK, but running OCCT isn't?RegardsBob ScottColonel, USAF (ret)ATP IMEL Gulfstream II-III-IV-VColorado Springs, CO

Don't quote me Bob, but my interpretation of what I read was that NVIDA will throttle on any detection of overcurrent. I am not sure what AMD did.