Hi,
 
I was wondering what people have been able to reach on they 970 FTW cards, i've been playing a bit with mine yesterday and was doing with
+125 mhz
+400 memory
Score: 2371
Temp: ~71c
Which was written in uningine valley benchmark like 1636 clock and 3900 memory if i'm not mistaken (not changing TDP).
I will post the link to the benchmark tonight if i can.
Seem like a nice chip, how is your's doing?
Quite happy with mine at the moment just need a backplate and a block!
Anamnesis

Here's an interesting thing I'm experiencing, I left the stock settings on the 970 FTW, and tried playing Call of Duty Advanced Warfare on all settings maxed at 1080p.  My GPU temp hit a whopping 75C.  Any thoughts on that?  I'm currently stress testing a variety of games to see how the temperature stacks up.  But in general is running at 75C often ok for the card?

shatel86
Here's an interesting thing I'm experiencing, I left the stock settings on the 970 FTW, and tried playing Call of Duty Advanced Warfare on all settings maxed at 1080p.  My GPU temp hit a whopping 75C.  Any thoughts on that?  I'm currently stress testing a variety of games to see how the temperature stacks up.  But in general is running at 75C often ok for the card?

Mine did the same thing until I used Precision X16 and set a fan profile, after that I see 68-70C usually but don't have COD:AW to test with.

Mine can only do +85 core, +190 memory. Max I see is 1510mhz, but it cant even maintain that in games, it just clocks down to 1430-1480 usually depending on the game. Wish there was a bios out that would give me consistent performance.

Please share your temperatures using default gpu/fan settings (stock so no precisionX)
 
Metro 2033:            78c   (1920x1080 everything maxed)
Valley benchmark:   76c   (Preset Extreme HD, 1920x1080 8xAA fullscreen)

Man I can barely get +100.  I'll crash in Valley Benchmark after 20 seconds at +120.
The noise to cooling trade off I have come to accept is to have the fans spin up enough to keep the GPU at 69-70 C during full load.  Doing this, according to GPU-Z, my fans are spinning at 75%.
 
Here is the custom fan curve I came up with:

 
edit: Oh and OP, the FPS counter in Unigine Valley is highly inaccurate.  I think it said mine was at 1631 too, lol...

Not to mention that I have two 970 FTWs with two different core clocks under boost (at OEM settings/no OC).
 
Can't recall the degree of MHz difference between the two at present, but will follow up later.
 
In summary, yes, stating clock speeds would be helpful to all.

Should I turn K boost on??

Hennesey
Should I turn K boost on??

I can't say for sure but Kboost is really for likes of stress testing and benchmarking, turning it on just makes the clock speeds max out.
why do you think you should turn it on?
if you think you need it on ALL of the time, then no i do not recommend turning it on. or you will be sitting on the desktop and your gpu will be at 1400mhz or what ever YOUR clock speeds are.
But if you are wondering if you need to turn it on for benchmarking and testing overclocks, then it is up to you. Remember Kboost does not work if you have vga (Analog) display connected. for this reason I did not have Kboost on and I still managed to get mine to a high overclock and stable.
hope this helps :)

Chrcoluk,
Since you haven't included your +voltage information, I am going to take a shot in the dark and assume you haven't added any voltage yet (the +mV slider on the right side of precisionx 16). If that's the case, I'm going to first advise you to google around or just browse the evga forums here for Overclocking tips to get the hang of it.
It works as such: higher GPU clock speeds (MHz) require more voltage to run stably. You can/should start by incrementally increasing your GPU clock offset by small margins (10MHz) and stress testing after each increment for no less than a minute. If it doesn't crash (black screen and report of driver failure) you can continue to raise it. When it eventually crashes you then learned your card's limit for that amount of voltage (you might get +50MHz with stock voltage).
You can now either dial down the clock speed until it's stable, or you can increase your Voltage and start the process all over again.
Final notes, You should be watching your core temps throughout this entire process. The main limiting factor on any OC is temperature; adding voltage adds heat! 87+ C is dangerous territory! So, don't add tons of voltage (20+mV) without first knowing it won't fry your card.
Once you're satisfied with your clock speeds and temp (70-80C isn't bad) run a very thorough stress-test for anywhere from 20-45 min.
There's a lot more to the Precisionx tool that will help you OC so read up on that too!
*Sorry for the wall of text, I thought it'd be brief but I wasn't about to under-inform an upcoming OC'er

Now, as for insight into my own OC with this card (FTW+). So far, I've managed:
1512 core boost / 1320 core base (+105)
4775 mem (+635)
1.127V (max +87mV)
@ 30C idle / 68C max recorded - with stock FTW+ fans in an air-cooled case (HAFX10).

I'm pretty satisfied, for the minimal amount of tweaking I've done so far. I know I can squeeze a little more performance out of this card without breaking the Precisionx power limit (110%). The memory, especially, has more headroom to play with but I found some sort of throttling came into play between the power limiter and increasing the memory speed. The throttle limited me to 1496 [core] in most cases, and sometimes lower. I might come back to it for more tweaking, later, (the measly 110% power limiter is the real bottleneck on this card) but I really don't think I'll need to! My ol' Asus GTX 570 dcu2 could barely OC at all. I'm lovin this card!

I have the FTW and I was asking if the FTW+ which its better components fixes the issue.
 
I didnt increase my voltage only the power limit.  Thanks for the tips.
 
By the way since adding the backplate I do have lower temps, idle goes down to 20-23C now (with 20% fan), and with 45% fan at full load my temps are under 60C.  Previously it topped 70C.

i was tinkering with it yesterday, manager to get+120 on the core which was borderline stable and +200 on memory which was again borderline stable,
 
When using uniengine heaven 4.0 benchmark it stutters alot (even at stock) should this be happening or is their a problem with my cards or pc?

AsimY
i was tinkering with it yesterday, manager to get+120 on the core which was borderline stable and +200 on memory which was again borderline stable,
 
When using uniengine heaven 4.0 benchmark it stutters alot (even at stock) should this be happening or is their a problem with my cards or pc?

That should not be happening at stock speeds. Ask EVGA what to do. They will accept RMA if there is a problem.

Anyone running sli ftw setup? I am having trouble getting past 100 core 200 mem