So now my +300 memory overclock will crash maybe once every few days with an artifact (the squares) everywhere. 
 
I know overclocking is never guaranteed (obviously) but isn't not even being able to hold a +300 memory overclock just really poor? I get most of my OC fps increases from memory overclocks on this 1080ti and my other. I feel slightly disappointed now :( 
 
Never had a card that can't hold at least 500, I was really surprised when this was crashing at that point too, but now even at 300, :/ 
 
Anyone else been this unlucky with the Hybrid ftw3 cards?

Sajin
Gotta love that silicon lottery.

Pretty painful really after I waited so long for it, but I guess it is what it is lol.

Cool GTX
These cards are already OC
 
What OC you getting on the GPU ?
 
If you reduce GPU OC, does RAM stability increase ?
 
What are your Temps ? room and the diff parts of card(s)
 
Colder is always better, but louder - test with all fans at Max
 
What is running in background ?
 
How sure are you your MB OC are stable ?
 
Test with No -- OC on MB / CPU do not run XMP on RAM
 
You OC the PCIe lanes ?
 
You OC your Display ?

Temps are fine. All of them are below 47c. Memory Power GPU 
 
Fans are at 70% nonstop and as I said always cold.
 
It is the memory, has nothing to do with XMP, CPU, RAM, PCI-E lanes, or my display. 
 
I have two 1080tis as I said, and it only happens on this one and I'm surprised that it overclocks so poorly on the memory, thats all, so poorly that it feels unusual. 

AHowes
His card is stuck at 1.043v and believes it's holding his oc back.. Maybe it is on the ram.

Though no doubt if he drops the oc from 2100 the ram will go up if the card is starving for power.

My OC is not at 2100, if you read the old thread, I removed all of those OC's and it has absolutely nothing to do with it - as I said above, I'm just surprised that it OC's so poorly on the memory, even if I remove the coreclock from +75 (2063) to 0, it still crashes at +300 memory. 

Tear it down and inspect factory thermal pads ?

Well then go with rma since the card can't do 1.093v and roll the dice again.

AHowes
Cooling the men more will help. I could not get 13k on the men before water cooling like the other 2 ftw3 elite cards but after the ek block i get well over 13k.

Well the fans are 70% constant, it's strange.
 
Like, +300 memory crashed earlier, now I'm testing with +600 memory and it's fine. 
 
But the other day, +600 memory crashed and artifact-ed reasonably fast - I've never had this experiencing OC'ing memory ever, and by this I mean; if +300 crashed after a while +600 would almost instantly crash 100% of the time. 

It feels 'wrong' as if something else is at play. I hope that makes sense.

Yeah sounds strange no doubt.. all I can think is something is up on the power delivery on that card or maybe it will perform better cold.. when it warms up it won't do the same?

Maybe shop for some better thermal pads to help with the cooling.

Personally when I'm benching or even gaming on air it's 100% fans at 50c. There cranking at 75% at 40c right when it starting anything 3d.

Yeah watch the sensors and see which chips get hotter.. bet it's the ones between the gpu chip and power.

There's better pads out there.. just gotta find out which size (prob ask evga). Put a dot of thermal goop on the chips before slaping the pads on it for even better heat transfer.

Normally I do that when install the water block but for some reason the instructions from ek didn't mention it so I didn't bother :/ next time I will.

My (2) EVGA 1080Ti FTW3 Hybrids are Rocking for me so far - FOLDING 24x7 -- it is Much more critical than gaming - Not Error tolerant at all.  Have not tried 600 MHz on RAM yet.
 
My PC is stable
 

 edit: typo
 
 
 
 
 

Thanks,
 
That is 3 GPUs with no gaps or the temps would be better
 
Room is 8 C
 
Its unheated to keep things running Fast, Nvidia Boost 3.0 --> sure does love a Cold GPU

Cool GTX
Thanks,
 
That is 3 GPUs with no gaps or the temps would be better
 
Room is 8 C
 
Its unheated to keep things running Fast, Nvidia Boost 3.0 --> sure does love a Cold GPU

Wouldn't +500 on your mem be 6000mhz? not 5508?
 
I can get 2113mhz or something in games but crashes a while after, how did you get two cards that are stable 

demon09
 
when folding or similar loads like mining memory runs at a slower clock 

Force P2 power state off in Nvidia profile inspector to fix this.