If i put -k zec zec it doesn’t work… but if i put -k zec it works… any idea? 
Cpu I7 2900K
16gb Ram
3x rx 470 4 gb
Do you use risers for your GPUs?
I can only run -k Zec as well (well to keep things stable).
-k zec zec is the default which works but I see a lot more 0S/s
-k zec zec zec crashes instantly
Using RX480 8GB
3.4Ghz i7
32GB RAM
SSD Drive
I think trying to convert PCI 16 to PCI 1 is too much for the board imo - using an OEM motherboard.
09:47:44 pool : received new job #00000036288624a2
09:47:44 pool : not accepted: null (21)
09:47:44 pool : Retrying in 5 seconds…
09:47:47 gpu#0: submitting solution
zec-sa#0write: The file handle supplied is not valid
09:47:47 gpu#0: Mining thread quit
09:47:47 gpu#0: submitting solution
zec-sa#0write: The file handle supplied is not valid
09:47:47 gpu#0: Mining thread quit
09:47:48 main : zec-sa#0: 34.4S/s total: 34.4S/s
09:47:48 gpu#0: submitting solution
zec-sa#0write: The file handle supplied is not valid
09:47:48 gpu#0: Mining thread quit
09:47:49 pool : Connecting to equihash.usa.nicehash.com:3357
the 0/0sol/s issue happens when the miner gets a new job from the pool, but is in the middle of submitting a solution already, so it continues to submit an invalid solution instead of updating to the new job, and causes miner threads to quit and 0/0sol issue. @Genoil am i correct or am i confused? thanks for your hard work
For 0.6 I’m using a simpler script and the new watchdog (-f 2) I posted it above. been working well so far
1 Like
here is full log
zconn exception in: read_until
09:47:38 pool : Retrying in 5 seconds…
09:47:43 main : zec-sa#0: 33.0S/s total: 33.0S/s
09:47:43 pool : Connecting to equihash.usa.nicehash.com:3357
09:47:43 pool : Connected!
09:47:43 pool : subscribed
09:47:43 gpu#0: submitting solution
09:47:43 pool : authorized worker 157sqh1G3k5KRFMDG4WvK5i1WJYm5QTXYu
09:47:43 pool : received new target: 0x01e1e1e1e00000…
09:47:44 pool : received new job #00000036288624a2
09:47:44 pool : not accepted: null (21)
09:47:44 pool : Retrying in 5 seconds…
09:47:47 gpu#0: submitting solution
zec-sa#0write: The file handle supplied is not valid
09:47:47 gpu#0: Mining thread quit
09:47:47 gpu#0: submitting solution
zec-sa#0write: The file handle supplied is not valid
09:47:47 gpu#0: Mining thread quit
09:47:48 main : zec-sa#0: 34.4S/s total: 34.4S/s
09:47:48 gpu#0: submitting solution
zec-sa#0write: The file handle supplied is not valid
09:47:48 gpu#0: Mining thread quit
09:47:49 pool : Connecting to equihash.usa.nicehash.com:3357
09:47:49 pool : Connected!
09:47:49 pool : subscribed
09:47:49 pool : authorized worker 157sqh1G3k5KRFMDG4WvK5i1WJYm5QTXYu
09:47:49 pool : received new target: 0x01e1e1e1e00000…
09:47:49 pool : received new job #0000003628862940
09:47:53 main : 0.0S/s detected! Attempting to restart gpu#0
09:47:53 main : 0.0S/s detected! Attempting to restart gpu#0
09:47:53 main : zec-sa#0: 6.0S/s total: 6.0S/s
09:47:58 main : 0.0S/s detected! Attempting to restart gpu#0
09:47:58 main : zec-sa#0: 0.0S/s total: 0.0S/s
09:48:03 main : zec-sa#0: 0.0S/s total: 0.0S/s
09:48:08 main : zec-sa#0: 0.0S/s total: 0.0S/s
so it disconnects from the pool cuz of whatever read_until error is, when that happens the GPU need to stop and wait for a new job and the 0/0SOL/s issue should be solved cuz they keep submitting shares even while disconnected causing threads to die and 0/0sol/s
either this or when it disconnects from the pool it sent a solution before my worker was authorized, causing null(21)
when this happens gpus need to wait for an authorized worker before restarting work or 0/0 happens also
Thanks for your answer and time. Do you mean this script:
:loop
genoil.exe -c equihash.eu.nicehash.com:3357 -u YourBTC.YourWORKER -p x -f 2
goto loop
Do I just have to create a mining.bat with this script and it should work fine (I have 4 GPU´s) ?
I am really sorry for asking you, but I am at the beginning of learning mining and coding and I would really appreciate your help! Thanks in advance.
I dont really care what everyone thinks about coinsforall.io but moved there yesterday because of the instability of the miners on windows.
Running his latest version with the new silentarmy kernel on ubuntu and getting a total of 40 more sols for the whole rig then anywhere else.
The pool also pays by the way as I got paid many times there so far. I have to admit that its not problem free though.
Sadly some of my gpu’s cant handle 0.6 I get 0sols after a while with zec zec while 0.5 was stable all night.
I used that but it give me the same speed.
I have the same issue, how can his RX 470 4GB be faster then my 8GB, does it have something to do with os? Or bios modifications?
Which 8gb rx 470 do you have?
I have been told that MSIs have a low memory rate than the others.
I have an msi 8 gb and the max i can get from that is around 28mh in eth and 40 sols in zcash.
EDIT: with a modded bios.
@Genoil - can the sensitivity of the watcher be changed?
e.g. to allow 3 instances of 0S/s before restarting:
The miner stopped immediately and on restart the old .exe is still running:
I’ve had to created Genoil kill switch for this issue:
HardExit.bat
killtask /f /im "genoil.exe" /t
1 Like
MSI and Sapphire
Sapphire doing 30 stock = 725w = 75C
MSI doing 25 underclocked = 580w and 450w @ -0.98v = 60C
I am getting over 2000Sol/s with my Radeon 6990. However it keeps submitting incorrect shares. Here is one of the cores. I downgraded to AMD 13.12 drivers. Anyway to fix this?
zec-sa#0: 966.5S/s total: 966.5S/s
zconn not accepted: low difficulty share of 0.062499046325683594 (23)
zec-sa#0 submitting solution
zconn not accepted: low difficulty share of 0.062499046325683594 (23)
Overclocked it too much?
Yeah thats what i thought.
MSI memory speed is around 6700 mhz. Sapphires are 8000mhz.
You can apply 1750 timing to both 1875 and 2000.
Increase the mem to 2000 as well. Then you will get more hashes. It must be done through bios though. Afterburner did not change my mem clock at all.
Im sure someone can correct me if im wrong.
Ok I guess 0.6 was rushed a bit too early. I’ll remove the watchdog and rebrand it as 0.5.1
You can still run it like 0.5 with -z 0
Yes that’s the one. If you have more than one GPU you may need to play with the -P and -g switches so recommend you get it working on its own first and then use the batch file
--gpus,-g <x y z ...> set gpu(s) -g 0 1 2 3 ...
--plat,-P <n> set OpenCL platform id -P 0