GTX 980 Classified SLI Overclocking vs 970 FTW+ SLI

Not sure if you're still checking this thread but i'll respond anyway. The reason your 980 isn't reaching quite as high clocks as the 970s is because weaker chips by their nature are able to hit higher clocks. You'll notice people talking about and read in reviews about CPU's that an Intel I7 5960X 8 core monster cpu has trouble hitting 4.6ghz in some cases even under water, while an I7 4820K quad core or even the 5820K/4930K 6 cores can hit 4.8ghz at least around half of the time depending on silicon quality/lottery. This same concept applies to GPU's, the more "cores" and whatnot that a processing chip has, the more limited it's overall overclock headroom is. This is the reason why "high end" chips of an architecture typically have far lower clock speeds, for example look at the GTX680 which had lower power consumption, cooler temps, and had base clock speeds well over 1,000-1,100mhz in most cases. This is because the 680 was a "GK104" chip, meaning it's the "mid tier" for the Keplar (G-"K" for keplar) architecture. Then look at the GTX 780/780 TI and TITAN which are the "high tier" cards made with the GK110 Keplar chip, these are much more powerful, having huge amounts of Cuda cores, nearly 3,000 compared to ~2,000 on the 680; this is an effect of it having a larger more powerful GK110 which is the pinnacle of the Keplar architecture, but it also has the side effect of higher power consumption (780/780 use 250w tdp vs under 200 on 680), hotter temps, and a larger gpu die. Then compare the clock speeds, the 680 with it's ~2k cores has base clocks of over 1,100 while the 780/780ti/titan with it's ~3k cores only has base clock speeds of around 900. The reason is like i said, weaker chips (i.e. ones with less "cores") are capable of hitting higher clock speeds by nature because you don't have to overclock as many "pieces" if you see what i mean.

Most people could hit 1,550mhz on 970s with plenty saying they got to near 1600, but with 980's it's a tad lower. For example, look at Jayz2cents video on the EVGA superclocked, he hits 1,490mhz overclock with air cooling as his max; however, with the GTX980 classified he was able to push it to 1,580mhz. Note that the expensive 14 phase powered 980 classified only hits around the same as mid-high quality 970s, this is an example of that limitation. Now don't think that this means the 980 isn't much more powerful than the 970, because it is, especially a high clocked like the classified when you overclock it. If you have a 970 and 980 of equal brand/type, with equal clock speed; the 980 will ALWAYS thrash it because of the extra cuda cores and memory bandwidth etc..

Now, regarding ASIC, just ignore it pretty much. It's not a very accurate measure of a cards quality. Higher asic "sometimes" means better OC with air cooling while lower ASIC "sometimes" means better OC with water cooling but lower on air etc..

If you truly wish to push your 980 classified's, then yes, you have to use the classified tool and/or custom bios's. That's the thing that separates the classified from the FTW, i mean there's only a 13mhz difference between a 980 FTW and a 980 Classy's base boost clock speed. So in order to take advantage of that 14 phase power with dual 8 pin etc.. you have to bypass the TDP wattage limit and the voltage limit that is part of the Nvidia "greenlight" program which forces companies like EVGA to limit the max voltage you can use.

Once you activate the classy tool and some custom bios's you'll be able to bypass that voltage limit. Also, did you know that the 980 classified has three bios's on it? There's a switch on the card that has "normal" "OC" and "LN2" bios's, if you haven't already, try switching to the OC bios and see if it boosts your clocks. If not then you can try the LN2 or flash a custom bios to it and you'll be good. I've seen people hit over 1,600 easily on classifieds with water cooling and custom bios/classified tool.