Sometimes have the same problem go to UEFI and change pcie gen 1 to gen3 and restart everything work ok. The last driver is problem I instal 381.xx not new and the miner is not good tree or five restart you card work ok.
this wrong over clock setting dude
All of GPU power gose for memory clock and it`s useless with Zcash mining
just look at your Core clock it`s only 1594 .!!! no way you can get over 600 sols with that core
any way just link Power limit with Temp limit
and put power limit on 75-80 %
Core clock 180-190
memory clock -350 -380
i your core clock hit 1847 MHZ then you`ll see 700-720 sol/s
@Yerr0r if I decreas the memory clock from +400 to -300 I get approx. 0.20 sols/w less.
I adore when people put this misinformation in writing when it is a proven fact that you will increase your H/s rate mining Zcash when you increase your memory clock along with the core clock.
Run a 1080ti with just core at +75 run for 5 minutes get your base average; then leaving the core at +75 increase memory to +550 and run for 5 minutes - your base average WILL be higher
2 Likes
I tried before to squash this myth that overclocking the memory does nothing but I was met with a whole lot of dismissal. That being said, the 1080 Tiās already have a buttload of memory and it doesnāt overclock as high/well as the lower models. But it doesnāt negate the fact that overclocking the memory does indeed increase your clock speed. I wish people would stop saying it doesnāt.
Lots of folks put down over locking numbers and very few give you any indication of how stable those settings are. Iād be curious to know, particularly for hose OCāing their memory to +500 and upā¦how long are your systems running before crashing? How long are your cards running before dying? Iām having a hard time believing that theyāre going to run for 1y+ with aggressive settings like that, but Iāll be the first to admit that I have no data to support this.
Furthermoreā¦I did a back of the envelope call the other day and basically discovered that, at current difficulty, squeezing an extra 200 H/s out of my cards would yield 0.5 Z per year. Iām all for optimization, but Iād rather up my probability of long card life when meager returns like this are at stack. Simply an opinion though.
PS I got a nice +20 H/s per card (x8) for a 100 MHz memclock speed up . In most cases, it does work. FWIW, my 1080 Tis showed no perceptible increase in H/s between +100 and +300.
ZCash does not need to overclock the memory is enough alone in my core 5x GTX 1080ti works 85% power 180+ core
Here are the settings Iām using. Very stable. I have a scheduled reboot on Sunday and Wednesday and a notification if the miner restarts EWBF. If EWBF errors more than 3 times it also auto reboots the machine and notifies me, but that rarely happens.
Once more read it well before be angry one me 
i only have 75% power limit so i need my coreclock to hit 1850 MHZ area if i overclock my memory to +500 and i already had +180 on core clock my gpu will crash and miner as well 
any way it`s Diff from GPU and another with this Previous setting that i mentioned up i get 710-730 sol/s and each Gpu take 180-200 Watt
Obviously itās the temp limit slider you set it at 65c so when the card reached 65c the card will throttle the power to control the temp limit set by you.
as i said it`s Diff from each GPU and another one you need to discover how to get 1850-1880 MHZ on Core clock with less power limit so you can save your gpu from higher Temp
as you can see here alot of overclcok setting
some one here share with me his overclock setting for same GPU msi duke his setting was
80-85 power limit
+200 Core clock
+500 memory clock
when i used this setting on my gpu EWBF miner crash all the time
all of miners prog can`t mine at all at the end my system crash as well
but this setting worked with him very well but with me it did not work even we have same GPU model
Iām having the same issue here - cards that should be identical are not in reality. Iām going to check the BIOS for each GPU to see if they have different versions flashed. Either way itās weird - does not seem to be related to power consumption or temps in my case.
Every rig has itās sweet spot, you just have to find it. As Yerr0r pointed out - one of my rigs with two gtx 1080tiās runs great at 80% power, +200 gpu core clock and +550 memory clock, fans at 70%. I built another 4 card rig, same motherboard, processor and GPUās (MSI gtx 1080ti Duke) and itās unstable at anything more than 80% power, +175 gpu core clock and +300 memory clock. Start with everything at 0, see if itās stable then go up in increments of 50 until it gets unstable then drop back some and see if it smooths out. Realistically if you want good temps and low electricity usage then you should look for the area around 750 sol/s
Good luck!
Exactly. You are only as strong as your weakest link. My issue as well, is you are only as strong as your highest ambient temp during the afternoon. I have to keep my clocks down a bit because during the summer it got so hot in my office in the afternoon. There are many limiting factors. I tend to average about 715 sol/sec on most of my 1080 tis, but I have one that will not stay above 670 when up to temp.
This is just how the silicon lottery goes people. The design and ship any cards that meet the MINIMUM design standard for the card. Some cards might greatly exceed those standards though. Just like a lottery, you donāt quite know what you will get.