Hello, i am new to mining, installed my first Zotac 3070 Dual Fan Rig and im having horrible issues with failed shares, ive tried tweaking owerclocks but to no succes, can someone help me, this is my overclocking parameters
Fan 75, Core -400, Mem 2400, Power 130 W, 87.56 Succes Rate more then 1500 failed actions in the past 48Hours, ive tried tweaking the parameters but still i get failed shares
temps for both 55C, and the second 48C
I’m also quite new here, but I think 2400 is too aggressive on memory OC.
On my 3070 FE i’m running core 1000, mem 1750, fan 55, power limit 115. It push 59MHs at 46° and 104 w (I’m fairly conservative in my settings 
)
I suggest to lower it to 2000 and then increase it by 50 units and let it run for some hours (maybe a whole day) each time, until you see again the invalid shares. At that point you revert to the previous setting, maybe tweaking it by some units more, and you are done with your config. It’s a work that will take some time but then you will have a stable worker with almost no worries.
Yes but i tried earlier with 1800mem and still i had invalid shares i dont really understand it, how many invalid shares do you get with your settings?
I’m using hiveos only since a week ago, before I was using windows.
I’ve had invalid shares the past night for the first time, about 10 for each gpu (i’m running a small rig with 3 cards), but I’ve not yet understood the cause (the cards were not overheating, maybe a spike in electric power?)
Before this episode I’ve never had an invalid share, both on hiveos and windows. And the OC setting I was using on windows were comparable with these.
With mem @ 1800 it is very strange to have invalid shares… Have you tried to mine with the same card on a different computer to check its behaviour? Have you tried some graphic benchmark on the card? Maybe it could also be a defective card…
I actually havent and as you were writing i lowered the mem again to 1700 and in a couple of minutes there came 5 invalid shares in both of them could it be the drivers or do i need to update something in hive os or maybe change the pool from hiveon to somewhere else?
I’m using the latest os version and 465.31 nvidia drivers.
You can try to update them, but, if nothing changes, I suggest also to try to run a benchmark on the card. If it freeze or has strange behaviours during the benchmark it could be defective.
The pool has no influence on rejected share (they depends on hardware), only on stale (due to high latency in connection)
ill try tu upgrade the hive os first and then try the drivers if itll help anything cause for both of them to be defective i dont understand it i got them new!:S
its really weird cause after i updated the HiveOS now with 1500mhz its showing 57.29MHS while earlier it showed 52MHZ a couple of minutes have passed and no errors are happening on either of them ?
Which miner are you using? I used phoenix miner and I had same problem.
After I switched to NBminer, I am having one or two invalid shares per month for full rig.
I have 4 of those zotac dual fan white cards…
Well, changing miner is also an option… in these cases it’s only a matter of trying until a solution is found.
I’m using T-Rex miner in my rig (2070 Super, 2080 TI, 3070)
As said before, I’ve had a spike of rejected shares some nights ago.
I’ve tried to reroute al the cables in my rig, to keep the riser usb cables away from the power cables (they were running very close each other before). Since then no more rejected.
Maybe there were some interferences from the power cables on the data cables.
If you have a similar situation, try to find a better way to route your cables.
Hope this helps you.
It is not about the oc but it’s about the miner. Use T-rex (Nvidia cards) for a proper mining job.
use gminer. over.
1st possibility, i did was shielding my wifi usb dongle from my rig with alumimuim foil and perhaps get a USB extension for your wifi usb dongle. (this is due to the EM emmision from the rig/cards, fyi i ran 4 3080s)
2nd possibility is to use LAN instead of wifi, this improves dramatically.
This topic was automatically closed 416 days after the last reply. New replies are no longer allowed.
