Stock values to anything are there for a reason and sometimes this reason is well documented in a white paper provided by the design team. Now this design team holds onto knowledge why this value is there holding onto very expensive,complicated testing methologies behind it. As example your car's engine has a piston wall of certain X .mm because the acclaimed/certified designer put this piston in a CAD simulation under certain variables in controled enviroment for X amount of turns then he represented this in real world scenarios and extensive tests so he finally figured WHY this X number needs to be used so your engine does not die (or if it dies he is the one to control when and how aka planned obsolescence.) This is the same with our scenario its just a VGA (an overbuilt one) not a car engine (relevant). There are almost always 4 ways to use a product but this extra robustness enables some more headroom to take advantage of. 1. This product gets modded to work above its specifications providing more performance because it handle it but you get less lifespan vs stock. How much less needs time, samples and experience to figure. (Or just OC it 30% then 2 years later and after a poor reflow some poor guy buys it on eBay listed as "office work never O/C." = I hate this.) 2. This product stays at default specs providing default performance but respecting the factory values and holding on that extra robustness/built quality you are allowed to use this product for a longer period of time. 3. The product gets modded to work under its specs so it underperforms but it sometimes worth the trade (mining comes to mind in our example). 4. Combination of the above. (UV/OC in this case) Now I am into the no.2 mentality of the above but this card's cooling design does not allow it to work within specs. I have repasted this card 3 times with same paste (AS5) and having done tests from 2 of these times I got the same temps all around. I have pc built experience over 15 years and owned 3 heat intensive cards dual gpu vgas (ASUS ARES 2, EVGA GTX 295 CO OP, NVIDIA 590) so my mounting and repasting skills are out of question. This Vega is a kid's toy to me. Another thing (does not account that much though) is that my motherboard is oriented downwards so the Vega heatsink is actually pulled onto the actual chip helped by gravity a bit (yea this does not compare to the bracket clamping force but still +5newtons (around 500+gramms i guess)). On the card UV process or any process altering factory values, I would only do it if I could mod the BIOS but since it is digitally signed, any attempt to boot a vega card with the vbios modified will result in a failed checksum of which the vega gpu's integrated security processor will shut down the card. OverdriveNtool /powerplay methods are a workaround to me and their stability/effectiveness solely depends on the driver/OS and its services(.NET). Every serious modder having stability and effectiveness in mind would only trust BIOS modding heck even actual hardware hardmodding. Everything else is an inferior alternative workaround and this is known for ages. Now I am given a BIOS from factory that can be stable 100% on my card and draw considerably less power (30% vs 240w|20% vs 210w) having around 7% less performance. That both sovles my heat problems and is probably tested more that you or me or anyone can ever test. Thank for your recommendations but I would rather stay on this BIOS side solution till I upgrade my cooling solution. Kinda irrelevant but the Pro drivers may enable extra features that gaming ones lack. Most important ones is 10bit color support (not 10bit pixel format) system wide (though I do not have a 10bit monitor to actually test it) and also extensively tested drivers on professional software.