k|ngp|n
Led's are bright cuz u need to be able to monitor if the cards are dropping voltages through all the ln2 smoke :). In 4 way LN2, its almost impossible to see anything in that area when benching hard at times.

Vlada011
You are fine, GTX780Ti Classified is excellent card, but you probably didn't wish Kingpin so badly as me. 

VAL79
I'm so jelly, my life has resorted to refreshing the EVGA product page over and over from my phone every time I think of it.

JDookie
Okay, I'm going to have to say that I'm honestly impressed by this card. I just started overclocking today, and with 1.28V (which is 15mV BELOW default voltage) I'm 100% stable at 1266MHz on the core and 7800MHz on the memory. When I say "default" voltage I mean with the way the card ships from EVGA, which is with a single voltage boost switch set to the ON position, and no other voltage tweaking.

JDookie

ericc191
I just hope they are still putting higher binned chips on the regular Classified too. I'll see when mine ships, I guess lol.

Correct me if I'm wrong, but I didn't think regular Classified chips were ever binned.
 

cobz01
I have both, my 2 kingpins are 1241mhz stable and only +100 offset on memory with stock voltage.. My 2 ti classys do 1280mhz and +200 offset at 1.212v, but they do 1320mhz stable and +500 memory offset 8000mhz effective at 1.25v game and firestrike stable.. I still have to get them underwater though, in the process of building my loop..

Okay, we have a bit to discuss and clarify here.
 
First and foremost, when I quoted 1.28V that was taken with a multimeter, not what Precision or GPU-Z reported. If you go by what Precision reports, then my 1.28V would be 1.187V, which is wrong. The importance of a good digital multimeter (DMM) has never been more apparent to me than now due to the amounts of power these cards use, and also the major inaccuracy of the sensors on these cards. So, to reiterate, comparing any voltage without taking the reading with a DMM is not accurate, and it's safe to assume that whatever you are seeing in Precision, GPU-Z, or any other software program is a LOT lower than the voltage you are *actually* using. In the case of my Kingpin, the difference is 1.0V+. Precision was reporting 1.187V and my actual voltage was 1.28V.
 
So assuming you said 1.212V because that's what you set in Precision, your actual voltage could have been anywhere from 1.25 to 1.3V actual, and the 1.25V you quoted for the 1320MHz was probably in the neighborhood of 1.325 to 1.35V.
 
Next, the memory. Your Classifieds use Hynix memory, and the Kingpin uses Samsung memory. In addition to the brand memory being different, the Classifieds' memory has looser timings than the Kingpin's Samsung which is why they overclock higher. The thing is, a mildly overclocked Hynix produces the same performance as a non-overclocked Samsung due to the difference in timings. Now, Vince (K|NGP|N) has released an optional BIOS for the Kingpin with looser timings for those that just feel better by having their memory overclock higher. He has stated that it won't produce any more performance but did it purely to stop the crying.
 
Now, I have also owned regular Classifieds and the highest I could get my last one was 1437MHz stable in Firestrike, and with +600 offset on the memory but it required a massive amount of voltage, 1.58V to be exact, which brings me to my next point.
 
I started a thread a week or two ago comparing the Classified and the Kingpin clock for clock, and even though they were *almost* identical in every run, the Classified's scores were actually higher in every single benchmark, again clock for clock, but that was with both cards using their stock heatsinks. You see, you can't really compare these cards on air because they will, for all relative purposes, be exactly the same. The difference doesn't start to expose itself until you start pumping up the clocks, and of course the voltage. After my testing, the higher I went with the regular Classified the less efficient it became vs. the Kingpin which seems to just LOVE higher clocks and voltage. I'm not done testing this Kingpin on water, but so far it is maintaining it's efficiency for a much longer duration than the regular Classified.
 
So you see, it's not just the fact that one card can clock higher than the other, it's the fact that they Kingpin can *usually* do it with less voltage, and in my case, with a LOT less voltage.
 
Get those cards on water, and get yourself a good digital multimeter, and then you'll be ready to see some crazy numbers.
 
Almost forgot. A note on benchmark stability. For some reason, I am finding Valley to be much more stressful to my cards than Firestrike. I can run Firestrike almost an entire 100MHz higher than I can Valley. For example, at stock volts, I can run Firestrike at 1340MHz all day long, but if I switch to Valley, it crashes almost instantly. Just wanted to mention this for when you're doing your testing.