Could Afterburner AutoOC damage my card?

[nicob]

Hello!

First of all, sorry for the dumb question.
I have a Palit GTX 1060, unfortunately it has a really small heatsink but I'm completely satisfied with its performance. I never really wanted that extra overclock in every day usage, I was just curious and wanted to see how much Afterburner can get out of the card.
I used the Auto-Overclock Scan-Test feature in Afterburner. I put the power limit to the maximum (116%) but I wanted to keep it safe so I did not allow additional voltage control and I kept the maximum throttle point at default (82C).
During the test, the card peaked at 81C for a very little time once but for the most part it was 79-80C. But I've noticed that in the GPU-Z tab, the tab showed a maximum TDP at a peak of 121% (so +5% compared to the power limit allowed by the original BIOS) which was around 145W.
My question is, how could this happen? I thought power-limit were hard-set in BIOS. My other question is, could this do any damage to my card? I set everything back to default, and my card is working properly, but since then I've noticed that even at 80% TDP, in less hardware extensive games, my card constantly hits voltage limit during games. (almost constantly). The boost clock (at least the peak) seem to be around the same. Should I be worried about this?

Thank you

 

[MrPotatoHead]

it will hit it's max voltage limit as set by the bios when you're playing games, nothing wrong with that. It could also well be that GPU-Z isn't getting the TDP power draw from an actual sensor in the GPU, more trying to guesstimate what the power draw is based on the cards voltage, clocks and usage hence the slightly higher TDP power draw that GPU-Z is reporting when it should be hard locked in the bios. If it still plays games fine on stock settings and doesn't overheat, throttle or lower boost clocks then you should have nothing to worry about, also there's very little chance that you have damaged the GPU simply by raising the clocks slightly and the power limit as that is hard coded into the bios to stop permanent damage from overclocking.

 

[micropage7]

Should be not worried, except you find something unusual like higher temp than usuall or glitch

 

[nicob]

Thank you for the responses!
As for the temperatures, sadly they were always quiet high. In extensive games it stays around 78-79C and the fan can go up until 70% even with the default profile.
Honestly its rare to see that a 3rd party card has a worse cooler than the reference/founders edition but I should not complain, it works and never actually reaches that 82C spot.

 

[FreedomEclipse]

If you repaste the card with some good thermal paste, temperatures could drop another 5-8c if not even more dependant on how good the airflow is inside your case.

Even if it's something like MX-4 it will still be a lot better than their stock thermal paste.

 

[nicob]

Thank you for your suggestion but I did that twice already and although the original paste seemed very dry, overall it did not help with the temperatures.
Unfortunately there is nothing to do, the heatsink is extremely small, it does not even go all the way to cover the VRM and the pipes are not in direct contact with the GPU either. On this video you can see how small it is compared to the JetStream.

The weird thing is, there seem to be more revisions under the same name. Some Dual heatsinks are longer and cover the entire card. On mine, they did not even put thermal pads on the memory. Guessing fron the BIOS version, this seems to be a later revision. maybe the bicoin prices forced them to use even cheaper components. Either way, its funny to see people with their sub 65C 1060 everywhere on the internet. But it is working so I should not complain

 

[Final_Fighter]

i had a pny1060 6gb. its the same card as the palit duel but it had the heatsink that covered the entire card, phases, and pads on the memory. also had heatpipes that made contact with the die. ran pretty darn cool. i shunt modded it and ran it like that until the mining craze. sold it for $500 second hand on ebay at the peak of it all. bought it at best buy for 269(or was it 249?) either way i was pretty happy with that card. i did notice that pny, palit started making them cheaper as soon as the minig craze came around. the version you have has better flow over the power phases to prevent burnouts caused by minning as most minning software put extreme loads on the power phases that would not be present during any other type of workloads.

 

[FreedomEclipse]

Thank you for your suggestion but I did that twice already and although the original paste seemed very dry, overall it did not help with the temperatures.
Unfortunately there is nothing to do, the heatsink is extremely small, it does not even go all the way to cover the VRM and the pipes are not in direct contact with the GPU either. On this video you can see how small it is compared to the JetStream.

The weird thing is, there seem to be more revisions under the same name. Some Dual heatsinks are longer and cover the entire card. On mine, they did not even put thermal pads on the memory. Guessing fron the BIOS version, this seems to be a later revision. maybe the bicoin prices forced them to use even cheaper components. Either way, its funny to see people with their sub 65C 1060 everywhere on the internet. But it is working so I should not complain

At the same time there are different tiers. Cheap as shit cards and more expensive cards from the same range. Like the gaming OC range from gigabyte or the Gaming Armor range from MSI - Those represent the cheapest in a graphics card line up before you get to the better equipped gaming X, Z and Aorus cards etc. So it it depends on where you entered the market at.

Thats not to say those cards are bad though. I have a gaming OC card and I switched the cooler for a more heavy duty one and now it never runs above 53c and boosts to the same speed as every other 1080ti. Sometimes it's worth spending that extra money for an overall better card as far as the cooler is concerned