Ok i will try it tomorrow.
 
What should I do if I do not want to reduce the core clock? Increase volts? And how much can I increase the voltage without having to worry about?
 
best regards
Wurmi

Debating on stepping up to the FTW+... Not to sure if I should or not. Would cost me a total of $100 to step up as one of my FTW's I got for $334.99

Would it be better to step up or not really worth the time and effort?

Final video on this series: overclocking the 970 FTW+ and SLI benchmarks. I also share my opinion on the #RAMGATE issue and give some recommendations for folks interested in stepping-up or upgrading from different cards: http://youtu.be/8gWwoqrtmNQ
 
Enjoy. :)

jeffmd
Levifig, I watched your video and was intrigued by some things so I decided to test my ftw+ that I just got. I am not an OCer though so I won't be doing any of that.. I much rather run cool and safe. With that I will say I run the card with the dip switch for the fan in overclock mode to help run cooler. I find the second bios has a nice curve I can work with cooling vs noise. 
 
So right off the bat, gpu-z tells me my Asic value is %80. I am surprised though that this number is just out there.. where does it come from? how does gpuz calculate it? 
 
So using the gpu killer OCCT, at the start of the gpu test (which alot of people use to demonstrate the ram performance problem) the freq is all over the place, but it eventually settles in and on this test (I have no other test that pushes the gpu temp like this one) it settles at 1352mhz. I am not sure if this is good or bad, but since boost is usually supposed to be the temporary speed and the card is locked at 15mhz below it and staying there, that sounds pretty good. 
 
Even better however is that both unigin heaven and valley at maximum quality had the gpu locked at 1430mhz. 

Dude: you have an awesome card!! :D haha
So, about the tests: I noticed that tests like OCCT and EVGA's OC Scanner actually run at a slightly lower that target speeds. I'm not sure why, but I did notice that. Gaming benchmarks are a much more clear indication of OC than those synthetic tests.
 
A little note regarding my video (and I need to add a note to the video itself too) is that my 2nd card, the 69.2% ASIC one, actually runs at 1392Mhz out of the box. So it does seem to match up with the increase you're seeing in yours. It seems that anything under 63-65% will actually clock below target speeds, and above that it will target above target speeds, with the increase being proportional to the ASIC rating increase.
 
Having said that, nobody is really sure how that value is calculated (it seems to be hardcoded to the chip's metadata). Higher ASIC ratings don't even necessarily mean better overclockers (if you're doing phase/dry ice cooling you actually want lower ASIC ratings, as you're countering the excess leakage with your extreme cooling systems).
 
I'll let other weigh in on this but ya: out of the box, on air, higher ASIC usually means a more efficient card, thus performing better, especially if on stock cooler and stock clocks. :)
 
Thanks for the feedback. :D
 
 

sahafiec
great video levifig, thanks for sharing it with us.

Thank you! :)

diamond-optic
just changed the TIM on my 970 FTW+ to arctic silver ceramique  ...hoping for temps maybe 2-3C lower under load.
 
So far temps are actually like 1-3C higher both when idle and also while running OC Scanner, hopefully they go down after a bit of use (even though on previous cards that I did the same thing with, temps were lower right away) ...otherwise I just wasted my time :-P
 
Though after taking the heatsink off, I noticed there was bits of the stock TIM on random spots of the heatpipes and also a few smears on the mosfet plate lol. The contact plate also had a few easily noticeable scratches in it, they werent real deep but they still stood out.

Oh noes… :S I was hoping that would improve temps as well (marginally, I know) but this doesn't bode well for doing that… I guess EVGA did pick good TIM for their stock setup… ;)

Changed the TIM on my FTW+ two weeks ago. Got 1-2°C better temps (used the Gelid Solutions GC-Extreme).
 
Saturday I switched the cooler to a Raijintek Morpheus black. Could still use the front- and backplate, so it is an awesome "all black" setup now cooled directly with two Noctua NF-F12 industrial (2000 PWM).
Could connect the fans to the PWM header of the card to get them PWM controlled and even achieved to get the zero fan mode working.

Load
fans @ 1700rpm: 43-44°C
fans @ 1300rpm: 45°C
fans @ 1100rpm: 47°C
fans @ 650rpm: 55°C

Idle
fans @ 0rpm: 40°C (after load)

If the card used all of the vram, I would buy two. Guess I will have to buy 980"s

Has anyone found a water block for the FTW+ yet?

UserX
 
I am actually in the market for a Waterblock as well so If anyone knows of one please let us know.

 
All you have to do is scroll down a little in the thread to find this:  http://forums.evga.com/Waterblock-for-your-980-ti-FTW-I-found-one-finally-m2448388.aspx
 
Alphacool 980 M11
 
 

evga made in ?
 
taiwan or usa?