jaybee83

yeeeeeehaw! [emoji16]

Sent from my Nexus 5 using Tapatalk

As long as you don't use the cheapest you can find you should be okay with almost anything. I usually see people who have problems are usually using way too much or removed the heat sink and did not remove the old paste and reapply new. So many people see the paste there and assume it will be okay. Don't do that!!!

Also, new heat sink means a new tube of paste, dont try to reseal the tube and use on another install even if there is plenty left in the tube unless you are doing it on the same day.

say what now? are u seriously using a completely new tube of fresh paste for each application?!?! :-O

Sent from my Nexus 5 using Tapatalk

yep, could also be that ur gpu model was missing in the .inf

Sent from my Nexus 5 using Tapatalk

u have to disable driver signature enforcement first (google is ur friend)

Sent from my Nexus 5 using Tapatalk

he meant non 347

Sent from my Nexus 5 using Tapatalk

make sure to uninstall ur current driver first using display driver uninstaller (DDU) in safe mode. then reboot with driver signature enforcement disabled in win8/8.1 (google is ur friend) before installing mr. fox 344.75 driver.

Sent from my Nexus 5 using Tapatalk

unlikely. ive seen lots of user reports on those scratches caused by ICD, but ive never seen a case where hardware was damaged in a way that went beyond just pure visuals my previous 2960XM cpu also carried away scratches from countless ICD repastes over the years, but was running like a champ till the end

Sent from my Nexus 5 using Tapatalk

yep, thats also the only thing holding me back from getting those insane sticks [emoji14]

Sent from my Nexus 5 using Tapatalk

best thing is to properly test how much bandwidth the core needs at its given overclock. that way ull only be overclocking the vRAM as much as it "needs to be" thus, determine the max. stable core clock and vRAM clock, then determine at which clocks u get the highesr benchmark results

Sent from my Nexus 5 using Tapatalk

lol fox, your analogies are disturbing and amusing in equal parts

im assuming this is the 960M or 965M? this is just rotten...

in any case, ever more reason to invest more money into prema

Sent from my Nexus 5 using Tapatalk

nice, tonight ill sleep much more peacefully knowing this

Sent from my Nexus 5 using Tapatalk

ouch, that doesnt sound right... could also be due to those missing voltage regulators....lets see what meaker finds out, hes planning to add some extra ones to his card and see if it changes the oc ability

Sent from my Nexus 5 using Tapatalk

in that case, i think its better if u just send me those sammys, since it would be a shame to just clock them below 2400 [emoji14]

Sent from my Nexus 5 using Tapatalk

nvidia gpus have a builtin throttle mechanism for furmark/kombustor/occt, so i wouldnt use those to stress test ur gpu. just use a benchmark like firestrike, unigine valley or a demanding game like bf4 or crysis 3 instead

Sent from my Nexus 5 using Tapatalk

You might get a little out of it, but not as much as with Samsung sticks.

those samsung sticks are indeed insane! and so friggin cheap, already have a bad conscience when seeing all those unknowing sellers out there selling those sticks for dirt cheap just cuz theyre used *lol* just too bad they only come in 4GB flavour...

i just finished a one week tweaking marathon with my corsair vengeance sticks, you have NO idea how many times i had to reboot due to system crashes *lol* didnt take out the cmos battery even once

thats the good thing about intel xtu though: if u set unstable settings, then everything reverts back to default at the next reboot thats also why i prefer using xtu as compared to direct bios setting...

well in that case theres no way around premamod

intel xtu actually changes the settings in the system bios, but not sure if it would stick in linux as well. could be worth a try!

use intel xtu

easiest thing would be to install 344.75 instead

Sent from my Nexus 5 using Tapatalk

cooling aint the problem whatsoever (at least not on my rig), the gpu core itself will give out way before any thermal limits are reached whatsoever havent done proper stability testing yet, but ballpark for my 980M is +295 mhz on stock voltage and +440 mhz at 1.2v. thats +26% and +39%, respectively. so its not too utopic to predict my stable spot to be around 30-35% as for temps, ive barely just managed to break 70C under heavy load *lol*

Sent from my Nexus 5 using Tapatalk

so we "only" get like 30-40% overclock, what a shame....

ouch, sorry to hear that mate... doesnt have to be the speed per se though, maybe its just the specific brand thats incompatible...

good answer and another quality post done, just two more to go

especially for populating the secondary ram slots under the keyboard a service manual can indeed be very practical