Originally Posted by SSJVegeta
Would it be safe to use the 1.312v BIOS with my GTX 980 G1 Gaming that is air cooled?
It hits 50-55% fan speed and 75c with your 1.28v BIOS with 1582/4070 clocks.
I want to see how much more I can get out of this card. ASIC is 70.4%.
I'm not going to recommend that you do, especially with that ASIC% (which isn't a bad
one). I posted those 1.312v BIOS' because people kept asking for them (other threads were using 1.312v) and they MAY
come in handy for those with a really low ASIC% to squeeze a little bit more O/C headroom. I've said from the very beginning that I saw NO DIFFERENCE
from using 1.281v compared to using 1.312v with GM204
. Temps basically just went up and for me I felt I had less stability. I see other threads started backing down to 1.281v also so I really cannot recommend it. I leave it there for those that dig for it. If you wish to travel to the dark side be warned
! Something, something dark side...^ Mandatory homework for this thread...
Originally Posted by megax05
Thanks for the response man I did change the volt+PL sliders to max and i got identical reading in gpuz and msi OSD @ 1456 core, so I will start working from there and see.
Ok you're on the right track now. Raise GPU CORE +10Mhz at a time until you TDR (crash) and then back it down at least 10-20Mhz lower for stability YMMV. Let us know how it works
Originally Posted by OcSlave
Just wondering what you guys think to this...
Got a 2nd xtreme gaming today, swapped cards out and this new card is at least 10c hotter while benching Valley.
Other card can keep below 49c with 80% fan, this card is starting to push 63c with 80% fan.
For some reason my fingers are itching to replace the tim .
2nd card has a higher voltage and lower asic.
When oc'd 2nd card hits 1.243v, 1st does 1.193v and this is with voltage slider maxed
And i'm curious, maybe flash the higher voltage bios onto the 1st card so I have the same voltage and more for the 1st card or vice a versa for lower volts.
Running a corsair 850watt, Rise of tomb raider maxed out ram and cores has me pulling about 630 from the wall.
Womb Trader runs awesome at 4k
I class bottom card as 2nd card although afterburner says 1st ?
Replacing the factory thermal compound and heat pads IMO is always a good idea. Just need to do it RIGHT
. Do NOT use compound that is electrically conductive (like Arctic silver 5 is). And make sure you get the right thickness for thermal pads otherwise it can prevent things from sitting properly across the entire GPU.
With (2) GPUs in SLI
, your case will most likely need to be left open at all times. Clean up wires, dust, etc. you just doubled your heat management issues with a 2nd GPU
When you have (2) GPUs in SLI, even if they are using the "SLOT 1 & 3" combination where they are further apart, the TOP GPU is ALWAYS going to be considerably hotter under load. Heat from the backplate of the bottom GPU will be sucked right up into the TOP GPU.
I have also seen a 5C-10C difference at times although it was usually closer to more like 5C. This is perfectly "normal" but definitely not desired. Keep in mind that these Windforce cooler designs eject all the HOT air right back into the case and INTAKE all of that same hot air it is up to us to remove the hot air, bring in cool air and we need to be "forceful" to remove the heat at a high enough rate of speed (higher than it is being created).
When you MOD a GPU you are expecting to have to cool it a little better. A custom fan curve is pretty much MANDATORY.. and SUPPLEMENTAL cooling is also HIGHLY recommended. If I may re-post Chemical world's PICS, it shows an excellent job of someone who is tackling the GPU heat issue properly and beautifully. When you are dealing with GM200, HEAT will be your #1 challenge which is why water blocks are so desired for them.
Note the 80mm fan on TOP of the GPU backplate with tiewraps as little "feet" to allow air to escape? Makes a HUGE diff. His temps between the GPUs should be close to identical now under load.
As far as your voltages with the STOCK BIOS, this is completely normal and as you pointed out has to do with your ASIC% value. When using a MOD BIOS voltages are still automatically determined according to your ASIC% values in the lower CLKs but as you get up into the higher CLKs the voltage range becomes more and more narrow until CLK74 where the voltage is specified and no longer a variable. Voltage fluctuation under full load causes throttling. Not all GPUs are created equal and basically you are going to want to run BOTH GPU's at the same exact BOOST speeds under load and I would't worry about customizing the voltage for each GPU unless you have DRASTICALLY difference in ASIC% value between them. You are going to likely run into one of the GPUs maximum overclock ability and/or a heat restriction before you run into a voltage restriction when on AIR.Edited by Laithan - 2/13/16 at 9:24am