Asus GTX 780 TI - temperature (& other) issues

TIZ012

New member
Hi All,

Its been a while since I have posted in OC3D. I have a problem as I just got an Asus GTX 780 TI and I'm very worried about the high temperature readings from the VRMs. Software temp results seem to be different than the actual readings for this card.

Guru3d has done a review and used a temperature sensing camera to give actual temp results and not relying on software.

In other forums users have posted that resitting the heat sink on the VRMs and using a different thermal paste and buffer between head sink and component helps to reduce the heat by 10C...

Yet in other forums still, users have also posted that the VRM sensors are misplaced and the readings are from different components which are at a higher temperature than the VRM components (This sounds like excuses and BS to me though!)

I hope someone has some insight or knows about this or other issues with the Asus GTX 780 TI card.

All help is appreciated!

Thanks
 
ASUS GTX 780 TI OC version. But apparently most of the high end Asus cards have these high VRM temp issues?
 
I think he's more than likely referring to the GTX 780 Ti DirectCU II OC 3GB, or though it could be the Matrix or a reference card..
 
I see the problem I'm sorry I was referring to the DCUII cooler card...I though the only Asus gtx 780 ti OC is with the DCUII cooler on it...but i can see why it might refer to the matrix and so on...
 
Ok so I understand it as no-one has actually noticed the high vrm temperature the ASUS GTX780 TI DUII OC edition gets to? I'm hitting 90C easily with a core temp of 60C max. My case is a Cooler Master HAF 932, large case with lots of fans and good air flow...Any help will be appreciated in the matter. I am considering RMA'ing the card just because of this reason.

Thanks
Tiz
 
Ok so I understand it as no-one has actually noticed the high vrm temperature the ASUS GTX780 TI DUII OC edition gets to? I'm hitting 90C easily with a core temp of 60C max. My case is a Cooler Master HAF 932, large case with lots of fans and good air flow...Any help will be appreciated in the matter. I am considering RMA'ing the card just because of this reason.

Thanks
Tiz

Well you at least used the card. Ive just rma mine because on stress condition she was reseting my system....Did you purchased from Scan?
 
Hey Zero

No I was really unsure of where to purchase from. The only advantage of purchasing from Scan was to get a free game with it, which I already got for free from a friend who purchased two GTX 770s. I finally bought mine from Amazon which was only £5-£10 cheaper. But they have a 30 day no questions asked return policy. Of course as long as you haven't blown up the product (or thrown it out of a window etc.).
Also interesting you should say that, because for instance my GPU doesn't hang or break on any heavy benchmark....except windows performance index :D . Most of the time it completes the windows P.I. but occasionally it doesn't and like you said it just restarts the system. This is obviously not correct behavior of a high end half a grand card. I will most probably RMA it for another Asus. Which is why I'm still looking for a confirmation on the temperature issue.

Edit: I believe I am 90% sure now that I will be Returning my Asus and go with a different brand altogether, its strange after using Asus GPUs since my first build. I love the way the card looks and is built, but the temperature issue is the most frustrating thing of it all...also those few times that the GPU just randomly restarts the whole system isn't a sign I like either.

Tiz
 
Last edited:
Hey Zero

No I was really unsure of where to purchase from. The only advantage of purchasing from Scan was to get a free game with it, which I already got for free from a friend who purchased two GTX 770s. I finally bought mine from Amazon which was only £5-£10 cheaper. But they have a 30 day no questions asked return policy. Of course as long as you haven't blown up the product (or thrown it out of a window etc.).
Also interesting you should say that, because for instance my GPU doesn't hang or break on any heavy benchmark....except windows performance index :D . Most of the time it completes the windows P.I. but occasionally it doesn't and like you said it just restarts the system. This is obviously not correct behavior of a high end half a grand card. I will most probably RMA it for another Asus. Which is why I'm still looking for a confirmation on the temperature issue.

Edit: I believe I am 90% sure now that I will be Returning my Asus and go with a different brand altogether, its strange after using Asus GPUs since my first build. I love the way the card looks and is built, but the temperature issue is the most frustrating thing of it all...also those few times that the GPU just randomly restarts the whole system isn't a sign I like either.

Tiz

Ok i got mine today..and yep..under stress the vrm's temps hit 102 C....i really dont know what to do...im thinking anyway in the feature(very close) i will watercooling it...but question is....is just a missdesign of their cooler that vrm's are getting so hot..or is a mass fault?
 
Asus (reply to) VRM temperature

Hey 102 degrees C is hotter than what I get normally. Mine in the worst possible conditions, like an hour or more of benchmark tests, reach 99 degrees C. I also contacted Asus complaining a lot.

At first their response was just RMA the card...it felt like the guy didn't give a damn. Then I sent them another annoyed (u could say very anti-asus message) and the guy assisting me, finally seemed to take what I was saying into consideration, he asked me for GPU-z log files of my temperatures/or just a screenshot of my GPU-z information and temps...I gave him way more than he asked for.
In his response to my attachments, I got the feeling he didn't look through all of it, but I don't care because he wrote the magic sentence which I wanted to see from the beginning. "From the attachments it seems normal." what he says there shifts everything on Asus...now if I ever have a temperature issue with my card I have a record of my correspondence with this Asus official telling me its normal temperatures. After that I decided to keep the card, and like u I am considering watercooling it as well. But much further in the future as it costs a hell of a lot!.

Regards

Tiz
 
This days i was running some tests . I have upgraded my corsair graphite windowed 760 to enermax 14 cm fans
http://www.scan.co.uk/products/140m...ntrol-white-led-lightening-detachable-batwing

With this i was able to stabilize the temperature for asus very fine. They are not loud on 70% and also with asus fans at 70% the vrms temp was not going more then 75 C runing valey on 1100 Gpu and 7070 Memory. If i turn the Enermax to maximum will be some sound(the air sound only) and the card fans to 100%(i think they are at least 3 times louder then the enermax) the vrms are not going more then 58 C. And the enermax fans are looking kinda fancy :D

4g3jh5.jpg
 
hi guys,
(you may saw me on other forum regarding this issue)
what iv learned is:
these temps are within specs like asus say.
the build is very special due to the heat issue (14 layers of thick PCB)
high quality CAPS.
even though ASUS gave me some reasnoble answers... i still wasent able to sleep good at night.
iv got th CM 690 III and my air flow is super.
i also tried the Typhoon 12mm FAN 4250rpm on the 780Ti DC2OC....... dident even scrached its balls.
only solution is to use GPU TWEAK!
when I Use GPU tweak... the VRM never go above 92c (on extreme cases).

this is what iv done:

This is on the Default clocks:
http://easycaptures.com/fs/uploaded/773/8790639909.png

FAN:

http://easycaptures.com/fs/uploaded/773/2386247207.png


AND this is on the little OC: (same FAN as aabove)

http://easycaptures.com/fs/uploaded/773/6758552269.png
 
Day when I decide to leave ASUS graphic cards are on best decision I made.
I could only imagine problems if their cards work on some EVGA speed, example reference PCB with Matrix clocks, and similar things...
The best way is when someone want to buy ASUS in Serbia but wait OC model and wait weeks for that if no such model at the moment. I say best because that's usually something under 50MHz or under 40MHz, and then even post as signature OC version. Imagine their cards with +100, +150 +200 on GPU out of box. Imagine Strix OC with 104C on VRM on fabric 20MHz higher speed than reference with +120-130 on GPU out of box, how? Their support team no way to support such risk and gambling. They are good to connect controllers on motherboards and write BIOS. I hope they used 125C capable aluminium solid caps, because little friends near VRM are in deep problem. Boiling point under one of best cooler on stock speed.
 
Day when I decide to leave ASUS graphic cards are on best decision I made.
I could only imagine problems if their cards work on some EVGA speed, example reference PCB with Matrix clocks, and similar things...
The best way is when someone want to buy ASUS in Serbia but wait OC model and wait weeks for that if no such model at the moment. I say best because that's usually something under 50MHz or under 40MHz, and then even post as signature OC version. Imagine their cards with +100, +150 +200 on GPU out of box. Imagine Strix OC with 104C on VRM on fabric 20MHz higher speed than reference with +120-130 on GPU out of box, how? Their support team no way to support such risk and gambling. They are good to connect controllers on motherboards and write BIOS. I hope they used 125C capable aluminium solid caps, because little friends near VRM are in deep problem. Boiling point under one of best cooler on stock speed.

yes my friend, they do use all those high quality componant.
also their quality control is very theral.
 
Day when I decide to leave ASUS graphic cards are on best decision I made.
I could only imagine problems if their cards work on some EVGA speed, example reference PCB with Matrix clocks, and similar things...
The best way is when someone want to buy ASUS in Serbia but wait OC model and wait weeks for that if no such model at the moment. I say best because that's usually something under 50MHz or under 40MHz, and then even post as signature OC version. Imagine their cards with +100, +150 +200 on GPU out of box. Imagine Strix OC with 104C on VRM on fabric 20MHz higher speed than reference with +120-130 on GPU out of box, how? Their support team no way to support such risk and gambling. They are good to connect controllers on motherboards and write BIOS. I hope they used 125C capable aluminium solid caps, because little friends near VRM are in deep problem. Boiling point under one of best cooler on stock speed.

What core overclock are you getting with the kingpin, How are you finding it ?
 
I didn't overclock at all... I mean I try but on 1265MHz no more PT space, in 3DMark 11 on something under that, but something near 1265MHz is close and in Firestrike... On fabric 1200MHz 109% is power target because of that I use LN2 BIOS. I suspect 6300 GPU score was less because PT limit(default 5950-6050 depend of driver), I mean I didn't look all time boost but few times touch power limit, that mean maybe card drop clock on some place and need BIOS with disabled Boost. That's not nicest option for me, I would love some official BIOS with 130% PT, boost enabled and tight timings because Valley and other test pass 400 on memory with tight Samsung latency.
That's little under 2000MHz...For memory not need higher voltage, but I keep and that on default.
130% power target would be really fantastic for gamers, because then we can use to play on somewhere near 1300MHz, off course on water.
Voltage and BIOS I didn't change, I told you twice I had shut down on 1.35V on something little over 1350MHz on my previous GTX780 Classified. EVGA just send yesterday label for RMA my PSU and than I will install and hydrocopper and try something else. I was completely sure PSU was problem with 780 Classy because I warn guy who bought from me and his friend couldn't repeat that, because he don't OC specific but just to try.
 
Last edited:
I didn't overclock at all... I mean I try but on 1265MHz no more PT space, in 3DMark 11 on something under that, but something near 1265MHz is close and in Firestrike... On fabric 1200MHz 109% is power target because of that I use LN2 BIOS.
Voltage and BIOS I didn't change, I told you twice I had shut down on 1.35V on something little over 1350MHz on my previous GTX780 Classified. EVGA just send yesterday label for RMA my PSU and than I will install and hydrocopper and try something else. I was completely sure PSU was problem with 780 Classy because I warn guy who bought from me and his friend couldn't repeat that, because he don't OC specific but just to try.

You have never told me anything like that dude, Your confusing me with someone else, Nice clocks though.

Also maybe try a different brand of PSU and not just EVGA, Try Superflower if you can as they make the best and most stable PSU's currently on the market and are used by the worlds top overclockers or even Corsair or Silverstone PSU's, It's kind of silly sticking to the same brand for all the internal parts, Mix and match for the best package :)

I've also had EVGA PSU's and they fail quite a lot, Been through 2 of them, The quality just isn't that good when compared to other manufacturers.
 
Last edited:
Yes, I will use some of new generation G2/P2, no matter on price, and much cheaper unit is OK for me.
Important to be about 1000W.
That was nice project NEX1500, Etasis destroy that, obvious they didn't was ready for task. I can say I'm happy because PSU die but protect hardware from any damage even if I had very old graphic, older memory and older SSD in that time it's good because everything still work after that.
Now Super Flower is one of leaders, coil whine much less follow their units than Seasonic example. Sleeve cables are fantastic too, I will buy black model from EVGA Europe.
I know overclockers usually use AX series in most case.
But with new Platinum series EVGA is very close.
 
Back
Top