OK, the oscillating between 1430 and 1418 is due to throttling because you are probably hitting your Temperature Limit. You really should create a variable custom fan curve to increase fan speed according to temps rather than a steady speed. You should also slide your Power Limit all of the way to the right and click apply so it doesn't hold you back. So far your overclock is looking good. You should concentrate mostly on your Core Clock for best results before you start increasing your Memory Clock. There is only so much power available before you reach the TDP of the video board and the Core Clock will give you better performance increases than the Memory Clock speed will. Once you reach your maximum Core Clock then start increasing your Memory Clock. I have never run my Memory more than +250 MHz.

It can be mind boggling that is for sure. Nothing you can do using MSI Afterburner will damage your video board. The best way to see what is holding you back is to run GPU-Z in the background when you run your benchmark and see what the 'PerfCap' reason is after the run. https://www.techpowerup.com/gpuz/
It will be color coded so you can tell what caused the clocks to drop. It might be that there isn't sufficient load to require your video board to run the clocks that high.

Sajin
Sounds like your card still had room for improvement within the same power level. How is it possible? Simple... your card's gpu must be a really good efficient chip.

Yep, Silicone Lottery winner for sure.

himmatsj

Sajin
Sounds like your card still had room for improvement within the same power level. How is it possible? Simple... your card's gpu must be a really good efficient chip.

Is there a way to determine how much I can push within this power level (without contributing to a rise in temperature)? Is the a danger, if say I put a +300MHz clock and the GPU bursts?
 
I have a ASIC of 83% if that's what you mean by GPU lottery.
 
I'm trying to search on the Internet but have found no answers as to why this is the case for me. Everywhere, it mentions overclock will lead to rise in temperatures.
 
I hope somebody from EVGA can help explain why is the full potential of the card within the same power/voltage/temperature settings no unlocked from the get go.

Yeah, just keep upping the clock until the card crashes. The card has a temp limit so it's going try and keep the card at 80c no matter how far you push the clocks. Putting +300 would most likely just cause the gpu to crash. Crashing because of high clocks can be fixed by resetting precision x.

I have had 7 evga cards in the 900 series in my possession and none of them had an asic quality higher than 78. SO grats there on the lottery. It could be that you just have a great chip and when running at the higher frequency it is happy.
 
the only way to see what you can get too at what temp and frequency is to test it. i would push it up till it crashes or you see temps you arent happy with then dial it back a bit say -25. test again and use the MV slider to bring it stable if needed. just for reference my 980ti's in SLI run totally stable at +135 core +25Mv which nets 1506 with the power slider all they way to 110% and priority on temp. This is watercooled but if you adjust your fan speed and its not thermal throttling then I be happy.   

From what I can see so far you should be able to easily reach +100 with room to spare. Maybe even +150?

It wasn't that long ago that many of us, including myself, had the same questions and concerns. We will do our best to help you understand anything you need help with. Remember, there are no stupid questions, only stupid answers. So, anything you want to know just ask and we will be happy to help.

Higher temps can be compensated for using the fan profile. Unstable behavior only happens when you push your clock rates too far. This is how you find your board's limit. After it becomes unstable just back it down 10 to 20 MHz and you are good. Nothing you can do using Afterburner, including boosting the voltage, will damage your board. There are voltage and power limits built into your VBIOS that will insure the longevity of your video board by keeping you from exceeding the preset limits. Shortening the lifespan of the video board is a risk that you take when you 'mod' or modify your VBIOS to unlock the preset limits. This is a risky procedure and is usually only done by 'power users' who want every little bit of performance available. The minor overclocking that you are doing now will not harm your video board. The benefits are smoother gameplay with higher Frames per Second. You have been blessed with a very nice GPU chip and it is up to you whether you wish to take advantage of the extra performance available to you. I never overclocked my video boards until I got my first EVGA GTX-680 and never modded a VBIOS until I got my MSI GTX-980 Gaming 4G boards I am using now. I have watercooling so I don't need to worry about temperatures. I rarely go over 45°C even when gaming for an hour or more. I can understand your concern with this being your very first video board purchase. If you are happy with your gaming results then you don't have to OC. It is just free performance for the taking if you are willing to experiment with finding the maximum stable clock speeds.

I've been thinking about the Power consumption and temperature of GPUs and did some research and came up with a theory for this to satisfy myself. This may or may not be common knowledge, I don't know.
 
As we all know, P = V * I
 
Given that the voltage is being untouched, that means the V in the above equation is a constant.
 
That leaves I, the current. There appear to be a number of components on a GPU that could possibly utilize current. The major one obviously is the GPU core itself, and there appear to be others like the VPU, MCU and Bus.
 
A GPU that is already at 99% utilization for the core is already maxing out the current available to the core. A common understanding is that the GPU core will be the first to reach 99% utilization, and in turn it ends up being the bottleneck for other components on the GPU.
 
Now, when applying an overclock, the core clock speed is increased, which causes its performance to increase. What this means is that, the other components on the GPU (like the VPU, MCU and Bus mentioned above) are able to up their utilization a bit more. Maybe the MCU and Bus were at 70% utilization when the core clock speed is 1250MHz. When increasing the core clock to 1350MHz, the MCU and Bus could increase to 75% utilization.
 
This is a small increase, but based on what little observation I have done, it is these supplementary components on the GPU that contribute to an ever so slight increase in the current passing through the GPU. So, this leads to a slightly higher "I" value.
 
Of course, if the voltage values are increased, then the GPU temps will increase noticeably. But by only increasing the core clock without touching the voltage, one could theoretically increase the clock speeds by +100MHz on a 900-series card and see very little if at all any rise in temperature.