Sound, Overclocking and Power Consumption
Despite the nearly non-existent shroud, the two fans running under full load are surprisingly quiet on the EVGA GTX 1080 Ti SC2.
The EVGA GTX 1080 Ti idles at about the same level as our reference card does, which is to say nearly silent by our testing equipment. Under a full gaming load, even though its slightly overclocked, it only generates 38.6 dbA, lower than the Founders Edition, lower than the Titan X, and nearly to the level of the GTX 1080!
Click to Enlarge
Under load testing in Unigine Heaven we saw the EVGA GTX 1080 Ti SC2 clock speeds hover around the 1830 MHz mark, well over the 1671 MHz rated by the GPU Boost clock settings out of the factory. This is great, and shows that NVIDIA and its partners continue to under promise and over deliver when it comes to clock speed ratings. Interesting, looking at the orange line in the graph above for temperature, we did see the GPU hit ~75C before the fan speeds ramped up enough to settle in at the 70C mark that EVGA had intended. I’m not sure if we’ll see a fan curve adjustment in software for this, but it was repeated (to a lesser degree) in our OC testing.
As for overclocking, I will admit that there was less headroom in our testing sample than I had originally expected. I was only able to get a 50 MHz offset running stable, even with the power limit cranked up to its maximum of 120% and the voltage increase peaked as well. Still, that did get us close to that magical 2.0 GHz mark.
Click to Enlarge
Click to Enlarge
With the overclock in place in Heaven, temperatures settled in the 72-73C range while clock speed averaged almost 1950 MHz.
The 50 MHz offset, coupled with the power limit and voltage increases through EVGA Precision XOC software were able to net us another 111 MHz over stock settings – not too shabby for a card that launched with a 1557 MHz base clock!!
Power Consumption
Click to Enlarge
Click to Enlarge
In our testing, the EVGA GeForce GTX 1080 Ti SC2 doesn’t use any additional power over the reference version of the GTX 1080 Ti. We see real-world draw in the range of 250 watts in both Rise of the Tomb Raider and The Witcher 3 testing which is within the stated TDP of the product and under the 300 watts of theoretical power draw from the combination and power sources (motherboard, two external PCIe connections).










Why didn’t use Metro Last
Why didn’t use Metro Last Light for power consumption testing?
Because in Russia, power
Because in Russia, power tests your consumption.
No reason – this was just a
No reason – this was just a quick spot check of everything on stock performance. You can see the Metro LL results from our Founders Edition review: https://www.pcper.com/reviews/Graphics-Cards/NVIDIA-GeForce-GTX-1080-Ti-Review/Detailed-Power-Consumption-Testing
Based on the fact that RoTR and W3 showed idential results, I would expcect the EVGA GTX 1080 Ti SC2 to follow suit.
Interesting. Thanks for the
Interesting. Thanks for the review.
Thanks for the review. Just
Thanks for the review. Just like to point out 3 things:
1) The power target of the card is 120% which is lower than some other cards. That’s not a huge deal in the grand scheme of things but it does seperate cards from each other.
2) Often times I see too many positive reviews for AIB cards. If everything is good then everything is average. It’s always interesting to see which AIB cooler is better but that’s a test few people can run.
3) The SC is not reference PCB and does not work with full cover water blocks. Last I checked neither EK nor Watercool have plans to develop one for it. The SC Black is reference PCB and EK has plans to make one for FTW3.
To address the power target
To address the power target question above, you cannot use the Power Target Value (Percentage) to compare from vendor to vendor. The reason is the default (100%) value can vary depending on the VRM solution.
With that being said, the EVGA FTW3 version of the card has a bit more robust VRM solution and will allow for a higher power target.
Right. I agree with you and
Right. I agree with you and it would have been more accurate if I specified the wattage instead.
I have one of these, rocks a
I have one of these, rocks a very stable 1949 MHz without any additional overclock, have yet to see it break 70C. Very nice, worth the wait, thanks EVGA!
Great Review!
Not sure why,
Great Review!
Not sure why, but almost every other manufacturer is going 2.5slot thickness for their custom cooler designs, seems like EVGA should and hopefully will do the same.
If you plan on keeping the card a long time like I do, I’d wait for one of the EVGA 2.5slot (or maybe even 3.0 slot) designs like a GTX1080Ti Classified or KingPin variant.
And the custom colors EVGA is offering this season look amazing white black green blue and red.
I’d go for an EVGA 1080Ti KingPin in white or black : )
WHOA, the MSRP for the R9
WHOA, the MSRP for the R9 Fury X is still $650 ?
I got mine on Amazon for $299
I think the MSRP Fury X should be 10% less then the GTX 1070 at most.
So in the $340 range.. but then I have a feeling AMD stopped production long ago, and at $320 its not making any money.
Anyways, at $300 I love that card.