Jump to content
Search In
  • More options...
Find results that contain...
Find results in...
Daymin

Drop your Temperature by 10°C

Recommended Posts

Posted · Original PosterOP

I've got a good deal on a used MSI Aero GTX1080.
The card does have a blower style cooler so it can get really hot and therefore loud under load.

To reduce noise I wanted to try changing the thermal paste with some thermal grizzly stuff I had lying around from my EK waterblock.

 

Below is my before and after benchmark which speak for itself I think.
A 10°C difference with a time effort of 15 minutes and no cost.

 

Wanted to share my result as this a mod everybody is able to do.

 

 

Unbenannt.thumb.PNG.85928a90a93367c310199ad5ae3e1b68.PNG

Link to post
Share on other sites

if you want to improve even more at no cost and increase performance at the same time try to lower the power limit (i use 92% personally) and increase the core offset (i use +200) you'll get better performance than out of the box and you will have much lower temps as well.

 

i have a windforce card but my guess is that the difference will be even more significant with a blower card.

 

My card out of the box was running at around 76-77c under prolonged full load (unigine heaven) and with these settings max temp is now 70c under same condition. with fans running at 58% in both cases.


| CPU: Core i7-8700K @ 4.89ghz - 1.21v  Motherboard: Asus ROG STRIX Z370-E GAMING  CPU Cooler: Corsair H100i V2 |
| GPU: MSI GTX 1080Ti Gaming X Trio 2ghz OC  RAM: 32GB T-Force Delta RGB 3066mhz |
| Displays: Acer Predator XB270HU 1440p Gsync 144hz IPS Gaming monitor | Oculus Rift S

Link to post
Share on other sites
Posted · Original PosterOP
2 minutes ago, i_build_nanosuits said:

if you want to improve even more at no cost and increase performance at the same time try to lower the power limit (i use 92% personally) and increase the core offset (i use +200) you'll get better performance than out of the box and you will have much lower temps as well.

 

i have a windforce card but my guess is that the difference will be even more significant with a blower card.

 

My card out of the box was running at around 76-77c out of the box under prolonged full load (unigine heaven) and with these settings max temp is now 70c under same condition. with fans running at 58% in both cases.

That would be the software approach but that depends on how good your silicon is.
Additionally I do have a 120% power target to get the maximum performance.

Link to post
Share on other sites
8 minutes ago, Daymin said:

That would be the software approach but that depends on how good your silicon is.
Additionally I do have a 120% power target to get the maximum performance.

you are not getting any noticeable amount of extra performance by doing that...pascal trottle hard and quick with heat...all you are doing is pushing more heat, using more power, and trottling more...try it...lower your power limit and increase your core offset to compensate and test it, you'll see :)


| CPU: Core i7-8700K @ 4.89ghz - 1.21v  Motherboard: Asus ROG STRIX Z370-E GAMING  CPU Cooler: Corsair H100i V2 |
| GPU: MSI GTX 1080Ti Gaming X Trio 2ghz OC  RAM: 32GB T-Force Delta RGB 3066mhz |
| Displays: Acer Predator XB270HU 1440p Gsync 144hz IPS Gaming monitor | Oculus Rift S

Link to post
Share on other sites
17 minutes ago, i_build_nanosuits said:

you are not getting any noticeable amount of extra performance by doing that...pascal trottle hard and quick with heat...all you are doing is pushing more heat, using more power, and trottling more...try it...lower your power limit and increase your core offset to compensate and test it, you'll see :)

It depends on your cooling solution, yes in this case the limiting factor is the gpu's cooler, but my 1070 maxes out at 42 degrees C, and benifits from having a higher power limit and voltage limit. Actually in my case the limiting factor of my gpu performance is the power limit.

Link to post
Share on other sites
1 minute ago, Mavflight09 said:

It depends on your cooling solution, yes in this case the limiting factor is the gpu's cooler, but my 1070 maxes out at 42 degrees C, and benifits from having a higher power limit and voltage limit. Actually in my case the limiting factor of my gpu performance is the power limit.

yeah if you have an hybrid or a watercooled card and you don't care about pulling extra power from the wall then you are absolutely correct...but pascal i believe hits the first step down in clocks at around 60c already and it goes fast from there...so pretty much any air cooled card you can get 98% of the max performance by using a lower power limit and pushing up the core clock offset instead...it stays quieter, cooler and it still boost very well.


| CPU: Core i7-8700K @ 4.89ghz - 1.21v  Motherboard: Asus ROG STRIX Z370-E GAMING  CPU Cooler: Corsair H100i V2 |
| GPU: MSI GTX 1080Ti Gaming X Trio 2ghz OC  RAM: 32GB T-Force Delta RGB 3066mhz |
| Displays: Acer Predator XB270HU 1440p Gsync 144hz IPS Gaming monitor | Oculus Rift S

Link to post
Share on other sites
Posted · Original PosterOP
14 minutes ago, i_build_nanosuits said:

yeah if you have an hybrid or a watercooled card and you don't care about pulling extra power from the wall then you are absolutely correct...but pascal i believe hits the first step down in clocks at around 60c already and it goes fast from there...so pretty much any air cooled card you can get 98% of the max performance by using a lower power limit and pushing up the core clock offset instead...it stays quieter, cooler and it still boost very well.

I think you do have a small error in your logic.

 

You do need more power to achieve higher clocks.

Higher clocks = higher temperature

higher power limit = higher temperature

The fact that you can decrease your power limit and increase your clock is a proof that you do had luck in the silicon lottery.
That won't work for everybody.

Link to post
Share on other sites
Posted · Original PosterOP

I do have a 4k display so I can max out my graphics card in anygame.

30 Watts more from the wall unter full load doesn't really matter.

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


×