Well, what do you know, back to the early 2000, where soldering would get you a quadro from a geforce of a fifth of the price 
http://www.eevblog.com/forum/projects/hacking-nvidia-cards-into-their-professional-counterparts/
The 690 turned into a full k5000 with the straps hard re-configured.
It’s how the first hacks ages ago (talking 2000-2002) were done before soft modding the drivers to ignore the hardware Id came about.
The only difference between quadro and geforce has always been mostly that, in different guises over time, but never one of hardware that would truly make quadros a superior product, just something they banked in on at a premium to provide drivers with a different set of priorities in mind.
nice
now we just need to find out about the geforce 660, 670, and 680’s - if they have the exact same sort of switch or not.
wow, this is very cool… so they weren’t crippling drivers, just masking things with straps.
interesting.
They will probably use an ever so slightly different patching, but it will be the same mechanism if they have a quadro equivalent.
If even the PCD and upper layer map are the same, then chances are it’s exactly the same patching array, just set to different values.
They are, but whether the drivers are crippled/configured a way or another depends on the ID the videocard reports, and that ID is set by that small array of resistors.
Several people already did apparently. Me? I’m not taking a soldering iron to my videocards. Done it in the past, talking past when geforce model numbers were single digit 
but the resistors were thrice the size and the PCB had half the layers.
It’s not as hard or risky as it seems actually if you know what you’re doing, and that array only changes the ID, you could remove them all, run the videocard, and put them back on, and it’d work perfectly fine (you’re not overvolting anything or affecting the powermanagement or other similarly dangerous things).
You do need the right soldering iron though ,and a relatively steady hand.
The back of my GTX 670 looks different than the 670 speculation picture later in the thread.
I might be willing to try it on my 670 if and only if someone posted a step by step process on the 670 specifically. Otherwise I’m not going to experiment with my card until it’s been proven to work on the 670.
It can be slightly different not only for each model, but also for each manufacturer.
The nVIDIA manufacturing specs aren’t publicly available, but I doubt the ID array has any specific guidelines on WHERE it should be placed, as long as it’s trailed right and reports the right number.
You can figure out the settings yourself with a tester and/or the quadro equivalent to know what ID you’re supposed to return and how it’s obtained, but the empyrical method, while relatively safe, might take you a long time.
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.
