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

Questions about my 780s

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

Leifus

New Member
Joined
Jan 7, 2016
Hello, i am just wondering about some things about my Evga Gefroce GTX 780 SC ACX SLI setup.

Question 1:

I have been trying to overclock my 780s for some time now and then i ofc have to look at the temps. Then i noticed that my main card is getting up to 80°C when under max load(when not overclocked) at 1100Mhz. I am keeping my sidepanel off for now, to not hit them 80s in temps just to be safe. So is there something i can do to keep the temps low on my card? if i have my sidepanel of i will get alot of dust in my case.

Thermal Compound:

so i went so far to change thermal compound on my chip and under you can see the resualt.

Original thermal compound:
idle temp: 50°C
under stress: 80°C
Under stress while side panel off: 71°C

Cooler Master E1 IC Essential thermal compound:
idle: 41°C
under stress: 80°C
Under stress while side panel off: 71°C

Airflow:

For my computer case I got Corsair 750D and in it i got 140mm fan in the rear of my case that are pushing air out also a 240mm radiator at the top with push pull that are also pushing air out, my inntake is in the front with 120mm x2 fans.

Question 2:

When i overclock my GPUs by +50Mhz clock speed and stress test it dosent crash while testing, but if i play games for a while, suddenly my nvidia driver crash and somtimes my image just freezes, but i can hear sound for people in Teamspeak. So is this that i am hitting the overclock cap or is it that the driver is failing? Driver version 361.43

Question 3:
Do anybody know what material the heatsink of the 780s i have are made of? i am planing to try out Coollaboratory Liquid Ultra thermal compound, but i can't use that on Aluminium cus then it will corrodes and destroy my heatsink/cooler

Spec:
OS: Windows 10 Pro 64-bit
PSU: Corsair RM 850W
Motherboard: Z170 Asus ROG MAXIMUS VIII RANGER
CPU: i7 6700k @ 4.5GHz CPU Cooler: H100i locked water cooler
RAM: 16GB 2133Mhz HyperX Fury DDR4
GPU: EVGA Geforce gtx 780 SC x2
SSD: Samsung EVO 840 250GB
HDD: 2TB and 1TB
Case: Corsair 750D
 
Last edited:
If you're crashing the Nvidia driver you've pushed the speed too high for that voltage.

Can you post a picture of the build? I want to see the cable management.

The heatsink is mostly aluminum.
 
Thanks for your answer.

If you're crashing the Nvidia driver you've pushed the speed too high for that voltage.
Why dosent it crash in the stress tests i do then?

Can you post a picture of the build? I want to see the cable management.
I can take a picture of my case when i get home and replay to the thread then(about 3 hours).

The heatsink is mostly aluminum.
well, then i will use it on my CPU only then.
 
Why dosent it crash in the stress tests i do then?
No stress test is perfect, some games are more intensive then the test you're using. Try using heaven benchmark and loop it a few times as a stress test.



well, then i will use it on my CPU only then.
Unless you're benchmarking the Cpu and looking for every last mhz there really is no reason to use Coollab liquid ultra. It's also a pain to remove, if you need to, down the road.

If your temps are dropping 9c with the side panel off the case, get more airflow through the case. Are you sure the fans are set up in the direction as you posted above?
 
Last edited:
That looks like a pretty good motherboard.

(1) I am assuming that you have the 2 graphics cards in the "gray" PCIe connectors? These are 60 mm apart and would let the cards breath more.

(2) If you have 1 in the bottom "gray" and 1 in the bottom "black" PCIe connector, you would only have 40 mm spacing between the cards...this would choke them for air flow.

If the case is (1), do you have a card in between the two graphics cards, or in the bottom "black" PCIe slot? Either of those would choke air flow to the cards.

A picture would tell us more! :)
 
so here is my picture, i dont think cable management is the problem.(got to love my pink wall btw)
4qmyoj.jpg
 
Why dosent it crash in the stress tests i do then?

I had much more difficulty overclocking my last SLI setup than I did the individual cards. I could get a stable setup during stress testing in Heaven, but have the driver crash at the Windows desktop when I was surfing the web. It was an art, but I had to increase both the voltage and the clock on the graphics cards differently than when I did them individually. For some reason, the "faster" card's voltage was not going high enough to support the overclock of the slower card. However, when you start tweaking with voltages, your cards may run hotter...so the temperature problem is the higher priority.

- - - Updated - - -

It looks like you have more airflow going out of your case then you do going into the case (3 fans out, 2 fans in)...and the lower intake fan is solidly blocked by the lower drive bay.

It looks like you can move the lower drive bay closer to the PSU...that may get more air into the case.

- - - Updated - - -

Also, if you move the lower drive bay closer to the PSU, it looks like you could add another intake fan on the bottom of the case.
 
It could be worthwhile to position a fan at the end of the GPU's to help air get between them, but cable management isn't your problem.
 
So as i get it now, what i can do is move the drive cage 1 step to the left and put a inntake fan under where it was before or under both?

I overclocked a bit too, i got +120MHz with 1200mV and dident crash in heaven or firestrike. i was trying out how much MHz i could have gotten if i maxed out the mV at 1300mV, but when i did that my main card didn't want to go above 800MHz, but my second card did. So do anyone know why this happens?
 
bump(Not to sure if i am allowed to bump here tho)
 
Last edited:
Sounds almost as if the top card was throttling because of temp/poerlimit combination. Back off the voltage as much as possible to maintain your OC.
I just wanted to point out as well that SLI and game crashes when otherwise stable could just be the driver. Running SLI can be a pain at times if the driver doesn't have an SLI profile for that game. Different drivers may help. Tings I have done in the past is, if you know the game GFX engine (eg. Cryengine) I have used NVidia inspectors driver tools and selected a profile from another game that uses the same engine. That quite often works.
 
It is deffently not my temps, cuz they are at 70s so they should be good, but that powerlimit thing i dont know. I am just want to see how many extra MHz i can get from my card at max voltage
 
Back