Just upgraded my Dual 980 Ti to Dual 980 Ti SC for FREE

[Berryracer]

So I just did a free upgrade for myself by flashing both my EVGA 980 Ti GPUs with the EVGA 980 Ti SC BIOS which only adds 100 MHz to the GPU core but I'd rather have this small and stable overclock done at the BIOS level because if I tried to overclock using EVGA Precision X, even though the benchmarks would be stable, when I watch videos I would sometimes get some flickering or the whole player would crash.


I was able to achieve +250 MHz overclock on the memory using EVGA Precision X but again I don't want any instability so I'll settle for this GPU overclock for now from 1000 MHz to 1102 MHz.


Score went up significantly though even though it's a very small OC


@Mr. Fox


GeForce GTX 980 Ti SLI (355.60) [W10]




GeForce GTX 980 Ti SC SLI (355.60) [W10]




GeForce GTX 980 Ti SLI (355.60) [W10]






GeForce GTX 980 Ti SC SLI (355.60) [W10]






http://www.3dmark.com/3dm11/10171917

 

[RampantAndroid]

Take a look at your ASIC quality in GPUz to get an idea of what the card should really be capable of.

 

[Berryracer]

Take a look at your ASIC quality in GPUz to get an idea of what the card should really be capable of.

Can you please tell me how?

 

[RaulF]

Open gpu-z
On the top bar where it says techpower up gpuz right click
Towards the bottom look for read asic quality.

 

[MentalIlness]

What would a Asic quality of 65.6% be?

 

[iiiankiii]

What about the noise level? How much did it increase? I'm pretty sure those stock blowers blow at anything besides stock settings. My titan x was loud a mofo before it went under water. 2x blowers sandwiched together might be loud to maintain those overclocked speed.

 

[guskline]

berryracer, I have only one EVGA GTX980TI SC (now under water with an EK block) and it's fast stock. Two of them must be incredible. WOW!

 

[guskline]

My ASIC on the 980TI SC is 72.1%

 

[RampantAndroid]

What would a Asic quality of 65.6% be?

I don't know mappings of ASIC quality to OC ability, but generally it's a indicator of how much you can get out of the card in terms of its relationship to voltage. 65.6% may be one of the lowest values I've seen someone post online.

Doesn't mean you can't OC, it means you need more voltage (and may hit absolute power limits faster) to get the same OC levels as someone who is at 75%. My 980 at 74.3% seems to be capable of 1506.5 core, 3800 RAM. I disabled boost via BIOS to force my overclock, since the driver itself doesn't want to boost to 1506.5 without more voltage...but I'm 100% stable with that boost when I force it. I'm pretty sure if I upped my power limit from 285W (which is already up from 250W) to 300W I'd be able to run at 1506.5 or higher even in furmark (right now, I throttle in furmark.) My card unlike most 980s (which have 2 6pin connectors) is capable of 375W of total power.

PCIe power limit: 75W off the board
6 pin connector: 75W
8pin connector: 150W

So stock 980 = 75W + 75W + 75W = 225W
MSI 980 = 75W + 150W + 150W = 375W.

 

[moonbogg]

I got 1 card with about 66 asic and the other card is like 78 and they both clock to over 1500mhz in SLI so I'm not sure how important that asic reading is.

 

[kami]

I got 1 card with about 66 asic and the other card is like 78 and they both clock to over 1500mhz in SLI so I'm not sure how important that asic reading is.

Supposedly higher asic = lower voltage to achieve the same clocks.

My 980ti is 72% and I can hold at 1500mhz with 1.20v (using modded bios). Haven't really tried pushing past that (could maybe lower the voltage a little more still), but I briefly tried 1.23v and was doing 1530mhz easily. Honestly there's so much performance here (I run 1440p) that anymore ocing is pointless. I just keep it in the 1400s usually.

 

[RampantAndroid]

Yup - ASIC quality being higher means less voltage/less power/less heat at the same frequencies compared to a chip with lower quality. Both might hit the same clocks of course.

 

[moonbogg]

How much does temps effect the power usage of the card? For instance would a 980ti at 84c use much more power than a 980ti at 50c with same clocks?

 

[RampantAndroid]

How much does temps effect the power usage of the card? For instance would a 980ti at 84c use much more power than a 980ti at 50c with same clocks?

The GPU is more or less a space heater. The card that's running at 50C, assuming all else is the SAME (heatsink thermal coefficient and such) is using ~40% less power.

Edit: Fixed my derp. 40%, not 60%. 50 is 60% of 84....so 40% less. Derp.

 

[Sabrewings]

The GPU is more or less a space heater. The card that's running at 50C, assuming all else is the SAME (heatsink thermal coefficient and such) is using ~40% less power.

Edit: Fixed my derp. 40%, not 60%. 50 is 60% of 84....so 40% less. Derp.

I think he's referring to different coolers being involved and the higher power usage of warmer silicon.

For example, a card running at 50C on water with the same voltage, clocks, and utilization of a card running at 84C with an air cooler will use less power. The warmer silicon gets the more power it starts leaking for the same conditions (which makes cooler performance issues even worse the warmer you get).

 

[RampantAndroid]

I think he's referring to different coolers being involved and the higher power usage of warmer silicon.

For example, a card running at 50C on water with the same voltage, clocks, and utilization of a card running at 84C with an air cooler will use less power. The warmer silicon gets the more power it starts leaking for the same conditions (which makes cooler performance issues even worse the warmer you get).

Ah, fair. Then yes, it's non-linear how much it leaks based on how hot it is.

Either way, I maintain what I said before - these things are pretty much space heaters...so I think my above stands. All else being the same, the amount of power being used would be 40% less (you don't just generate heat from nowhere.)

 

[PhIlLy ChEeSe]

Go hang out at the EVGA forums [redacted]! Oh yeah it wasnt for free, you just think it was! Look at the bill......................EVGA= you paid dearly for it!!! FREE LOL!

Use appropriate language for a technical forum. If I see it again, you aren't going to be using any language in this forum.
-- stahlhart

 

[RampantAndroid]

Go hang out at the EVGA forums [redacted]! Oh yeah it wasnt for free, you just think it was! Look at the bill......................EVGA= you paid dearly for it!!! FREE LOL!

What?

evga cards are more or less priced in line with the rest. He achieved clocks that he otherwise would have had to pay more for. He could probably get even higher.

 

[Berryracer]

What?

evga cards are more or less priced in line with the rest. He achieved clocks that he otherwise would have had to pay more for. He could probably get even higher.

Thank you sir, my point exactly. Got a small bump for free. :awe:

 

[Kenmitch]

Thank you sir, my point exactly. Got a small bump for free. :awe:

Well free is debatable....You did pull a couple hairs out trying to get the flash software to work.