boredgunner

Enstranged

Confirmed as well, reference GTX580 WILL go to 1.15V in MSI afterburner with this bios update! cheers!!

Hmm, you sure you don't have an SC card?  Or did this "Fan Unlocker" suddenly enable 1150 mV?  The GTX 580 has the same VRM as the GTX 480.  I'm gonna go all the way to California to EVGA's office and beg for moar voltage (1200 mV minimum).  Or I'll use another modded BIOS.  

 
I live about 28 miles from EVGA headquarters.  Would a picket sign out front satisfy everyone's yearning for protest?

boredgunner

Enstranged

Confirmed as well, reference GTX580 WILL go to 1.15V in MSI afterburner with this bios update! cheers!!

Hmm, you sure you don't have an SC card?  Or did this "Fan Unlocker" suddenly enable 1150 mV?  The GTX 580 has the same VRM as the GTX 480.  I'm gonna go all the way to California to EVGA's office and beg for moar voltage (1200 mV minimum).  Or I'll use another modded BIOS.  

 
 
Correct!  So we know what is possible, hehe

I was limited to 1.138V with my reference "stepped-up" card so I KNOW I have the reference.  Now I can push it to 1.15V in Msi.  Weird stuff but I'll take it!! My temps aren't too bad but I don't trust my 700w psu too much so still waiting for the AX to show up before I try to unleash 1ghz on air :p

Awesome! Been waiting for this..

If you are giving stress to your voltage of card, you may need a second power supply in order to get clean power to your VGA 

Does this not unlock the fan to run as 'auto' without running the tuner software? I hate running extra "crap" for no reason ontop of Nvidia drivers. If anything, I use the Nvtuner that is a separate download from Nvidia alongside the drivers to make custom overclocked profiles.
 
Will this not allow 30% there? If so.... useless for most really. 

codeman

Does this not unlock the fan to run as 'auto' without running the tuner software? I hate running extra "crap" for no reason ontop of Nvidia drivers. If anything, I use the Nvtuner that is a separate download from Nvidia alongside the drivers to make custom overclocked profiles.

Will this not allow 30% there? If so.... useless for most really. 

 
This is essentially a BIOS update, so no additional software is ran.

Big problem flash my gtx 580 OC in sli now shows core 50 shaders 101 and mem 135?  Help shows same in e-leet and cpuid.  Something wrong in these two programs? Does clock properly in EVGA Precision. I used file you provided and screws up info shows the above problem.

Just weird that showed 3d clocks until I used fan unlocker now shows 2d core clocks instead of 3d.  At least its normal.  Thanks Jacob for fast response.  I have backplates and high airflow brackets coming.

codeman

Will this not allow 30% there? If so.... useless for most really. 

 Given the misconception that there is a "default" fan speed,  your question is valid.  However, there is no default fan speed, there is a default profile, and the profile sets the fan speed as a function of temperature.  The default 40% to 85% profile has not been changed, so you will never see  30%.  The value of the update is that you can decrease the magnitude of the hissing sound of the fan if you so choose.  There's no logical reason to have any restrictions at all, but 30% is less bad than 40%.  I just hope that with the release of Kepler next year, EVGA will be proactive and fix the bios before they release their cards instead of waiting for people to complain.  This would set EVGA apart from all other brands and thus increase sales. 

oOfuZedOo

Doesn't work for me. Just telling me "No display adapters were found that are compatible with this update"

Running 2 GTX 580 BO in SLI

Can you take a GPU-Z screenshot?

You are running the update for 1582 right?

roller11
Given the misconception that there is a "default" fan speed,  your question is valid.  However, there is no default fan speed, there is a default profile, and the profile sets the fan speed as a function of temperature.  The default 40% to 85% profile has not been changed, so you will never see  30%.  The value of the update is that you can decrease the magnitude of the hissing sound of the fan if you so choose.  There's no logical reason to have any restrictions at all, but 30% is less bad than 40%.

Close, but not quite.  Specifically:
 

The default 40% to 85% profile has not been changed, so you will never see  30%.

 
That isn't true.  The default profile has been changed.  That is what this BIOS update addresses.  The video card BIOS does not have more than one fan profile in it.  The BIOS only has one profile programmed and therefore it is not possible to change the profile while also retaining the old profile.  It replaces the default profile with a new profile which allows for 30% to 100%.  Now, the new profile is similar to the original profile so that the fan speed at various temperatures will be similar to what they were before the update.  Simply speaking, if you plotted this profile on a piece of graph paper as well as the default profile, the two would be similar in regards to profile and ramp rate.  But this certainly does not mean that you will never see 30%.  If the card is cool enough, the fan will drop as low as 30%.  Now, this will probably never happen unless you are ducting air conditioned air straight into the video card or have a very low ambient air temperature for whatever other reason.  To say that it isn't possible to see 30% isn't true.  On top of that, you will see the fan go below 40% -- dropping down towards 30% -- whereas before the fan would have never gone below 40%.
 
I do see my fan speed drop down in the 35% range during extended 2D load periods.
 
Also:
 

There's no logical reason to have any restrictions at all, but 30% is less bad than 40%.

 
This is also not true.  There is a logical reason to have fan speed restrictions.  All PWM fan controllers have built-in restrictions.  The restriction is incorporated to prevent the fan from stalling.   Once a fan stalls, it usually takes a considerably higher duty cycle to make it unstall.
 
Let me make this analogy.  First, we must assume that EVGA has determined that the fan won't stall at 30% (otherwise they wouldn't have released this update).  Let's say that the update allowed for unrestricted fan operation.  Lets say that your video card is sitting at 2D load for an extended time and the fan speed drops to 15% and then the fan stalls.  No big deal, right?  (That's what you apparently think since you say there is no logical reason to have restrictions at all)  You assume that as the card heats up, the duty cycle will increase and that the fan will surely start spinning again by 30% since we know it won't stall at 30%.  But this isn't the case.  Once the fan stalls, it may not unstall until 60% (for analogy sake) fan speed.  This would put the user on edge since he would enter a 3D load, see the GPU temperature climb rapidly while the fan was still stalled, and then all the sudden hear the fan go from silent to quite loud, and then see the GPU temperature plummet, and then hear the video card fan become much more quiet again.  Do you think that this operation may cause the customer to think his video card was broken?  Therefore, restrictions exist. 
 
Don't reply saying "but the fan speed would never go as low as 15% due to GPU temperature".  Sure it could: if you decided that the GPU should idle at 80c while 2D idle, the fan speed may drop as low as 15%.

Ok, I think I know where the issue is.
 
For you guys who have a Black Ops card, and are not able to apply the update, send me a mail at jacobf@evga.com.

ty_ger07
The default profile has been changed. 
I do see my fan speed drop down in the 35% range during extended 2D load periods.

Right you are.  I assumed that because there is no reason to change the default profile, EVGA didn't.  The result is that the fan will generate even less noise.   Thanks for pointing out my error.

There's no logical reason to have any restrictions at all, but 30% is less bad than 40%.

 

 

This is also not true.   There is a logical reason to have fan speed restrictions.  All PWM fan controllers have built-in restrictions. The restriction is incorporated to prevent the fan from stalling.   Once a fan stalls, it usually takes a considerably higher duty cycle to make it unstall.

There are at least two obvious flaws in your logic.  First, like many others, you are under the delusion that 0% means 0 RPM.  The truth is, 0% means the fan operates at it's designed-in minimum RPM which can be anything.  There's an graph  of this on the TIgerdirect site (Golf 140mm).  To prove this, connect your PWM fan to the PWM controller on your mobo and set it to 0%.  The 580 fan spins at 950RPM.  If you were right, then every PWM fan would be rated over a range of  0 RPM to it's maximum RPM.  But if you look at the specs on PWM fans, the minimum speed is never zero. 

Let's say that the update allowed for unrestricted fan operation.  Lets say that your video card is sitting at 2D load for an extended time and the fan speed drops to 15% and then the fan stalls.  No big deal, right?

Exactly right.  No big deal because the GPU is cool.  The fan need not turn AT ALL if the GPU is cool.  

You assume that as the card heats up, the duty cycle will increase and that the fan will surely start spinning again by 30% since we know it won't stall at 30%. 

I assume nothing of the sort.  It won't start up again due to  built-in hysteresis of the profile.  You grossly underestimate me.

Once the fan stalls, it may not unstall until 60% (for analogy sake) fan speed.  This would put the user on edge since he would enter a 3D load,

Your second error.   97C is 'on the edge', 60% fan occurs at 65C, well within the safe operation envelope.

see the GPU temperature climb rapidly while the fan was still stalled, and then all the sudden hear the fan go from silent to quite loud, and then see the GPU temperature plummet, and then hear the video card fan become much more quiet again.  Do you think that this operation may cause the customer to think his video card was broken?  Therefore, restrictions exist.

Third mistake, you're assuming that the fan minimum would force a wild profile.  In fact, the fan minimum can exist seperate and apart from  the crazy profile you suggest.  But even if your example was remotely possible, (which I disproved) all you would have shown is end users are capable of bad judgement, no reflection on the card.  In other words, your scenario is technically impossible and even if it happened, it would have absolutely no negative impact on the card.  Therefore, no logical reason to have any restrictions on minimum fan speed. 

Don't reply saying "but the fan speed would never go as low as 15% due to GPU temperature".  Sure it could: if you decided that the GPU should idle at 80c while 2D idle, the fan speed may drop as low as 15%.

Sure it could, so what?  80C is well within the safe operational limits of the card, you're just reinforcing my point.