• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

1080ti icx power temps HIGH

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Rainwater

Member
Joined
Nov 10, 2016
Location
Portland
So I have been messing with GPU overclocks. These ti's do not move as quickly as the 1080 I had. I ran my 1080 ftw hybrid at almost 2.2ghz. Struggling to keep this ti stable at 2025 sustained. If I up the offset to about 82 it will stay at 2012mhz and my memory at 5999. But I just spent an hour or so playing the division and looked at all my temps with gpu z and the icx memory 3 and icx power 5 are scorching at max temp. Memory 3 had a max of 95c and power 5 had a max of 127c. This sounds extreme. Max gpu temp of 52c I have attatched a screenshot showing gpu z with the max temps displayed. It also shows my gpu oc settings. Voltage max, power max mem offset 496, gpu 80. These two temps really concern me. Is this normal? I don't really know what is actually being measured for power 5 but I don't believe it is safe having memory temps hit 95c. I never checked the temps at stock settings. I will do so and repost later. I noticed about a 15 fps increase from this overclock in the division.
Screenshot (8).png
 

EarthDog

Gulper Nozzle Co-Owner
Joined
Dec 15, 2008
Location
Buckeyes!
Doesnt precision show you exactly where/what sensor that is? I'd guess it's a vrm....assuming its accurate. 95C feels hot for vram...

Have you looked at any reviews of the card and see if they ran across this? What has eva said?
 
OP
Rainwater

Rainwater

Member
Joined
Nov 10, 2016
Location
Portland
I have not yet. This just happened this morning. Kind of just posted this on the fly taking off to run errands. I guess precision might huh. I will look deeper and see what I can find. I just figured maybe someone here over the last couple years has seen this before. I'm hoping it's just inaccurate. Literally all the other mem and pwr readings are abotu the same as one another except these two.
 
OP
Rainwater

Rainwater

Member
Joined
Nov 10, 2016
Location
Portland
So I have done a little research and there was a thread on the EVGA forums about someone having a similar issue but mem 2 and pwr4. EVGA had him rma it. I bought this second hand and don't suspect I will be getting any sort of warranty. I am going to open up the precision software and see if I can get it to track the temps in all areas. The post I just mentioned had pics of this software doing so. I have been wanting to order an ifix it protech tool kit anyway so I am doing that now with delivery tomorrow. I am going to tear this thing apart and replace all thermal pads. I am hoping there is a thermal pad issue. CAN SOMEONE PLEASE tell me what type and thickness of thermal pads to order? I am going to hopefully start teardown this weekend. Get my hands dirty and have some fun. :attn:
Will these work?
https://www.google.com/shopping/pro...FPVH70AUfTv1ddZfB3D4tju5HfETubreg,prmr:1,cs:1

So just an update, as I mentioned earlier I've been wanting the ifix it protech tool kit anyway so I went ahead and ordered that. And a 100mm x 100mm x 1.5mm thermal pad. And some Thermal Grizzly Conductonaut since I plan to delid this 8600k around the end of the month when i have a few days off. I will update if I proceed with the teardown, if I don't then whatever it's good to have thermal pads lying around.
 
Last edited:

EarthDog

Gulper Nozzle Co-Owner
Joined
Dec 15, 2008
Location
Buckeyes!
1. See if the card is under warranty.
2. If it isnt in warranty, ask evga for specs on the thermal pads.

Edit: or just order them and hope...lol..
 
OP
Rainwater

Rainwater

Member
Joined
Nov 10, 2016
Location
Portland
I am going to speak with them. I sent them a message and I am in the process of trying to get something figured out. I registered the card as soon as I got it. I remember there being a 10 series evga thermal pad issue and looked up a video from Gamers Nexus and I purchased the same size pads they had a link for. That's what made me think it could be a tp issue in the first place.
 

yoadknux

Member
Joined
May 6, 2016
Are you on air?
You don't have a lot more OC room. The max for a 1080ti is 2100 on water. Anything above that is not really stable. Either it requires too much voltage, which triggers power limit and causes clock fluctuations, or it just crashes. Typical clocks are 2038-2076 on 1.05V.
As for memory, newer batches could do ~6300, older batches, mine included, max out at 6000-6100.

I don't know if your temperatures are an issue. My card doesn't even have sensors for VRM and VRam. I expect an overheating VRam would artifact. Still, from a research point of view, would be good to swap the pads.
 
OP
Rainwater

Rainwater

Member
Joined
Nov 10, 2016
Location
Portland
I haven't swapped anything. And no I'm on water but it's the Evga hybrid. Overclock is pretty disappointing. I'm more like 2000-2038. Memory 6100. The temps on the precision xoc never get that high. It's just gpu-z. So I updated it t the newest version and it got way more out of control. Pwr 5- 255c, pwr 4-127c. So needless to say I don't any of it and I'm gaming just fine with no issues so I just don't think there is anything to truly worry about.
 

yoadknux

Member
Joined
May 6, 2016
I haven't swapped anything. And no I'm on water but it's the Evga hybrid. Overclock is pretty disappointing. I'm more like 2000-2038. Memory 6100. The temps on the precision xoc never get that high. It's just gpu-z. So I updated it t the newest version and it got way more out of control. Pwr 5- 255c, pwr 4-127c. So needless to say I don't any of it and I'm gaming just fine with no issues so I just don't think there is anything to truly worry about.
If you really want max OC, you can flash the EVGA FTW3 Kingpin bios. It doesn't magically give you better clocks, but it removes power and thermal throttle, so you can put it at 1.093V constant with no clock fluctuations. I had 2100/2088 stable with it. My card however was an Aorus and when you mix 1080ti's bioses of different brands some display outputs stop working (until you re-flash). That's why I went back to the stock bios that can do something like 1.05V 2076, 1.044V 2063, 1.025V 2050.
 

EarthDog

Gulper Nozzle Co-Owner
Joined
Dec 15, 2008
Location
Buckeyes!
I haven't swapped anything. And no I'm on water but it's the Evga hybrid. Overclock is pretty disappointing. I'm more like 2000-2038. Memory 6100. The temps on the precision xoc never get that high. It's just gpu-z. So I updated it t the newest version and it got way more out of control. Pwr 5- 255c, pwr 4-127c. So needless to say I don't any of it and I'm gaming just fine with no issues so I just don't think there is anything to truly worry about.
Have you tried EVGA Precision (the software for the card) yet?
 
OP
Rainwater

Rainwater

Member
Joined
Nov 10, 2016
Location
Portland
Yes sir, It doesn't seem to be able to allow me to track min/max temps but I messed around with it with the sensor window open and would hit the windows button and click on it real quick to check the temps and it never read higher than the 60's while at the same exact time gpu-z would show 255c and 127c LOL. So I am confident it's not reaching 255 Degrees Celsius. Or 127 for that matter. I should have been patient and investigated a little further before the panic thread. I just remembered the thermal pad issue back in '16 with the EVGA 10 series and jumped to conclusions. But if I remember correctly it did not affect the Hybrids as the vrm's are being cooled either by the block or the fan on the card. That is why I had originally returned my 1080 FTW after 2 weeks of purchasing it and buying the FTW Hybrid in '16. I will probably never use an air cooled card again now.