• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Compare 3 way SLI/CrossFire air temperatures

Gripen90

Gripen90

Gripen90

Associate
Joined
20 Nov 2009
Posts
2,050
Location
Haarby, DENMARK
I often find my self spending a lot time trying to find temperatures results for these kind of setups running on air, but I rarely find them since I know it's not many who are running such setups, so I thought a thread dedicated to multigpu setups of minimum 3 cards would be nice, so people can share what kind of temperatures they are getting and what they're doing to improve temperatures and airflow.

What are peoples setups ?
How's your fan setup ?
Hows your fan profiles ?
What temperatures are you getting ?
What have you done to improve cooling in the case ?

These are the kind of question and answers I'd like to share.

I will start with with a couple of setups I've had.

Year 2009.
Graphics setup: 2x MSI GTX 280 OC Edition 1GB & 1x Realtek GTX 280 1GB
Case: Cooler Master CM 690 nVidia Edition.
Case Cooling 2x 120mm 200rpm front intake fans, 1x 120mm 1850rpm in front of graphics cards. 1x 80mm bottom intake fan 2000rpm. 2x 120m 1800rpm top exhaust fans, 1x 120mm rear 1800rpm exhaust fan, 1x 80mm 2000rpm rear atx motherboard intake fan. 1x120mm intake side panel 1200rpm fan.
http://i95.photobucket.com/albums/l131/Gripen90/IMG_3378Large.jpg

Year 2010.
Graphics setup: 3x EVGA GTX 470 SuperClocked 1.25GB
Case: Cooler Master CM 690 nVidia Edition.
Case Cooling 2x 120mm 200rpm front intake fans, 1x 120mm 1850rpm in front of graphics cards. 1x 80mm bottom intake fan 2000rpm. 2x 120m 1800rpm top exhaust fans, 1x 120mm rear 1800rpm exhaust fan, 1x 80mm 2000rpm rear atx motherboard intake fan. 1x120mm intake side panel 1200rpm fan.
http://i95.photobucket.com/albums/l131/Gripen90/IMG_4809Large.jpg

Year 2011.
Case: Cooler Master CM 690 nVidia Edition.
Graphics setup: 2x Gainward GTX 580 1.5GB & 1x PNY GTX 580 XLR8 1.5GB
Case Cooling 2x 120mm 200rpm front intake fans, 1x 120mm 1850rpm in front of graphics cards, 1x 120mm 1350rpm intake fan in front of CPU cooler. 1x 80mm bottom intake fan 2000rpm. 2x 120m 1800rpm top exhaust fans, 1x 120mm rear 1800rpm exhaust fan, 1x 80mm 2000rpm rear atx motherboard intake fan. 1x120mm intake side panel 1200rpm fan.
http://i95.photobucket.com/albums/l131/Gripen90/PNY GTX 580/HPIM3843.jpg

Year 2012.
Graphics setup:
Graphics setup: 3x Club3D GTX 670 2GB
Case: Cooler Master CM 690 nVidia Edition.
Case Cooling 2x 120mm 200rpm front intake fans, 1x 120mm 1850rpm in front of graphics cards, 1x 120mm 1350rpm intake fan in front of CPU cooler. 1x 80mm bottom intake fan 2000rpm. 2x 120m 1800rpm top exhaust fans, 1x 120mm rear 1800rpm exhaust fan, 1x 80mm 2000rpm rear atx motherboard intake fan. 1x120mm intake side panel 1200rpm fan.
http://i95.photobucket.com/albums/l131/Gripen90/IMG_5327.jpg


Year 2013.
Graphics setup: 2x MSI GTX 780 3GB & 1x MSI GTX 780 TwinFrozr Gaming OC 3GB
Case: Cooler Master CM 690 nVidia Edition.
Case Cooling 2x 120mm 200rpm front intake fans, 1x 120mm 1850rpm in front of graphics cards, 1x 120mm 1350rpm intake fan in front of CPU cooler. 1x 80mm bottom intake fan 2000rpm. 2x 120m 1800rpm top exhaust fans, 1x 120mm rear 1800rpm exhaust fan, 1x 80mm 2000rpm rear atx motherboard intake fan. 1x120mm intake side panel 1200rpm fan.
http://i95.photobucket.com/albums/l131/Gripen90/IMG_5819Custom.jpg

The "fun" part is that regardless of setup my GPU temperatures have always been the same in load in the last 5 years - there may have been 1-2°c difference at the very max.

I've also tried using different fan setups with variations in RPM speeds but they haven't made an hughe differences. I've even tried 35mm wide 120mm 1200rpm fans on the side panel for more air movement towards the cards, but a 120mm 25mm at 1800rpm gave same results at lower decibel.

So as for now I made a little video show the kind of temperatures I'm getting with the current setup. You can fastforward to the end for temperatures.

Summery.
Idle temperatures:
MSI GTX 780 reference GPU 1: 35°c
MSI GTX 780 reference GPU 2: 33°c
MSI GTX 780 TwinFrozr GPU 3: 29°c

Load temperatures:
MSI GTX 780 reference GPU 1: 84°c
MSI GTX 780 reference GPU 2: 88°c
MSI GTX 780 TwinFrozr GPU 3: 80°c

I re-ran the test 3 times in a loop and the temperatures stayed the same.
The fan profile used is the standard nVidia driver setting profile.
 
Last edited:
Bought a Scythe Ultra Kaze 120/38mm 3000rpm fan I'll have to retool at bit to fit the side panel so it'll stay clear of the CPU cooler. I'll test if that will lower the temperatures on the Cards.
http://www.scythe-usa.com/product/acc/022/dfs1238_detail.html
f66ae9d9_Scythe_DFS123812H-3000_in_and_out.png
 
Last edited:
Got this sanded down on the top to fit snug under my CPU cooler when the sidepanel is mounted.
http://i95.photobucket.com/albums/l131/Gripen90/GTX 780/IMG_5843Custom.jpg

Going from a plastic Silverstone grill with a 120x25mm Cooler Master Sickleflow on the side.
http://i95.photobucket.com/albums/l131/Gripen90/IMG_5836Custom-1.jpg

To an open black fan metal grill with a 120x38mm Scythe Ultra Kaze.
http://i95.photobucket.com/albums/l131/Gripen90/GTX 780/IMG_5856Custom.jpg

I've connected it to my Zalman Fanmate 2 and running it at a voltage of around 10v at approx 2500~2600rpm.
http://i95.photobucket.com/albums/l131/Gripen90/GTX 780/IMG_5849Custom.jpg
 
Last edited:
Getting the Scythe Ultra Kaze has improved idle and load temperatures on all cards.That means their boost clock can be sustain higher and for a longer time.
A little video again. Noise of the Ultra Kaze and temperatures during Valley 1.0 benchmark.
 
Last edited:
Nice little thread and nice little collection of cards you've had :D

If i were to ever have tri-sli/tri-fire i'd have all the cards cooled by either a chiller or custom loop, never had more than 2 cards underwater and only have 3 cards on my open air test bench so never really had to worry about temps :)
 
Tonester0011 said:
Nice little thread and nice little collection of cards you've had :D

If i were to ever have tri-sli/tri-fire i'd have all the cards cooled by either a chiller or custom loop, never had more than 2 cards underwater and only have 3 cards on my open air test bench so never really had to worry about temps :)
Click to expand...

Thank you ;)

Yea I know, and that's why I made this thread :D because 90% of the people going for 3 cards are more are going to water cool them. So we don't see that many with air cooling an have something to compared with regarding temperatures and cooling setup.
I do understand people going for H2O cooling to improve overclockability, get better temperatures etc.
But I can only speak for myself that I don't trust water cooling systems at all neither the pre-build closed set nor the ones you put together yourself. While I think it looks great and understand the reasoning behind it, the risks involved are just not something that I'm willing to take :rolleyes:
 
Last edited:
I have quite a dent gap between my cards

2013-3.jpg


I'm ordering the optional BS-03A box though which will suck away the hot air and blow it out the back as the inside of my case is a like a furnace when gaming!!

BS-032.jpg
 
almighty15 said:
I have quite a dent gap between my cards

2013-3.jpg

I'm ordering the optional BS-03A box though which will suck away the hot air and blow it out the back as the inside of my case is a like a furnace when gaming!!

BS-032.jpg
Click to expand...

Looks like a good plan. If you have option for a side panel fan on the case you could make it an exhaust fan ?.
 
So nows the question ! has the switch in side panel fan had any effect on the overall performance of the graphics cards ?

I'll let the benchmark results speak for them self.

NB ! I will remind that the cards did NOT underclock to under the default 862MHz core at any point when they ran the hottest !

These following results are purely from the gain in performance due to GPU 2.0 boost being able to run for an extended period of time due to a cooler core !
Left row before with Cooler Master Sickleflow 120x25mm side panel fan. Right row now with Scythe Kaze Ultra 3000 120x38mm side panel fan.

3Dmark Performance test.
http://i95.photobucket.com/albums/l131/Gripen90/3dmperf.jpg
3Dmark Extreme test.
http://i95.photobucket.com/albums/l131/Gripen90/3dmext.jpg
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom