Topic: CCminer(SP-MOD) Modded NVIDIA Maxwell / Pascal kernels. - page 1111. (Read 2347601 times)

Thank you myagui. I downloaded KBomba's  miner control some time ago and didn't set it up yet. And my main concern is a backup pool right now. Looking at your info now and thx for your time helping me.

When I'm mining straight for BTC, I just use MinerControl:
http://cryptomining-blog.com/tag/miner-control/

It may take a while to fully configure, but I think that it is definitely worth it, for the ease of use (basically, set and forget). MinerControl takes care of switching between what is paying the most, and also switches when a pool is dead/offline.

When I'm mining a specific coin, I use a batch file with a single call to ccminer. I don't ever have a crash, so I never cared to make a looping system. This configuration does not include a backup pool though. I babysit my miner when I'm not using MinerControl.

Here's what I'm running right this instant:

@echo off
C:
cd "C:\1. Cudaminer\cudaminer-2014-02-28\x86\"
C:\Windows\System32\cmd.exe /C start "CACHE GPU-0" /HIGH /AFFINITY 0x5 "C:\1. Cudaminer\cudaminer-2014-02-28\x86\cudaminer.exe" -a scrypt-jane:CACH -o stratum+tcp://cach.catcoin.cz:3333 -u -p -i 0 -L 8 -l t64x4 -b 4096 -m 1 -q -d 0
C:\Windows\System32\cmd.exe /C start "CACHE GPU-1" /HIGH /AFFINITY 0xA "C:\1. Cudaminer\cudaminer-2014-02-28\x86\cudaminer.exe" -a scrypt-jane:CACH -o stratum+tcp://cach.catcoin.cz:3333 -u -p -i 0 -L 8 -l t64x4 -b 4096 -m 1 -q -d 1

This batch file runs 2 concurrent cudaminer instances - both get launched at the same time - as it is using a "start" parameter. The point of using 2 instances is to have a separate worker for each GPU, which improves the reported hashrate for this pool/algo. This launch type also let's you easily tweak priority/affinity settings, which is useful in certain instances (and specially if you are also CPU mining).
IIRC, I had to set the shortcut to such batch files with the "run as administrator" option enabled, otherwise priority and/or affinity would not stick.

Yes it's in the same folder and I'll try to understand what you just posted and give it a try. thank you myagui
Ps What do you use?

I'm assuming that your launch batch file is placed on the same folder as the ccminer executable.
For sake of example, let's say that ccminer.exe is stored in this folder:

c:\miners\ccminer\release99\

So your batch file would be right alongside ccminer.exe (placed on the same folder), but for completeness, you would call the executable with a full path string, like so:

:loop
c:\miners\ccminer\release99\ccminer.exe -r 1 -R 5 -a quark   -o stratum+tcp://anypool
c:\miners\ccminer\release99\ccminer.exe -r 1 -R 5 -a quark   -o stratum+tcp://anypool
goto loop

So now just replace c:\miners\ccminer\release99\ from this example with whatever is the actual location of your executable, and there you go. No clue this will help at all though, just a quick experiment that cannot hurt.

If you want to place the batch file in a location separate from the miner executable, you can just add a folder navigation command at the loop start, like so:

:loop
cd c:\miners\ccminer\release99\
c:\miners\ccminer\release99\ccminer.exe -r 1 -R 5 -a quark   -o stratum+tcp://anypool
c:\miners\ccminer\release99\ccminer.exe -r 1 -R 5 -a quark   -o stratum+tcp://anypool
goto loop

You might need to tweak the path strings depending on how long your folder names are, or if there's any spaces on them. Adding one more example in which one folder name is using a space - note the added quote symbols (")

:loop
cd "c:\miners\ccminer\latest release\"
"c:\miners\ccminer\latest release\ccminer.exe" -r 1 -R 5 -a quark   -o stratum+tcp://anypool
"c:\miners\ccminer\latest release\ccminer.exe" -r 1 -R 5 -a quark   -o stratum+tcp://anypool
goto loop

@myagui
I'm very amateur at this. Could you explain more on \ please? Thx

@tbearhere:

I'm not using loops myself and I don't have any crashes, but seeing as the error message is that 'ccminer.exe is not found', have you tried adding the full path to ccminer? Could be that the crash is somehow messing up the set working directory for the batch file. If you are already using the full path at launch, then the suggestion is moot.
The result would be something like:

:loop
\ccminer.exe -r 1 -R 5 -a quark   -o stratum+tcp://anypool
\ccminer.exe -r 1 -R 5 -a quark   -o stratum+tcp://anypool
goto loop


Good Luck!

Trying to figure out why ccminer doesn't crash on this computer but does on my rig when stratums go down and I have a 2nd pool backup and it crashes at the goto loop command . I just noticed that one card on this computer is the super clocked factory setting no oc needed is on the mobo and it stays at 1350 750ti sweet spot. So overclocking may be part of the problem. But it still shouldn't crash IMO.

:loop
ccminer.exe -r 1 -R 5 -a quark   -o stratum+tcp://anypool
ccminer.exe -r 1 -R 5 -a quark   -o stratum+tcp://anypool
And it crashes right here at the goto loop and says ccminer.exe can't be found. So it freezes right there.
goto loop

And if I click off the notification only it starts mining again without manually  restarting ccminer.

Actually the 750ti throttles at stock tdp (meaning when it reaches 100%tdp, it gets clocked down a bit to stay below the 100% tdp target...) increasing the tdp would allow it to be more stable regardless of the OC (also it is the voltage which needs to be increased for higher overclocking)

what settings do you have? ...

#crysx

wow - so much happened and i missed it all ...

was completely swamped today - and wont be back in the office till tomorrow now ...

the open mining rig - we started like that ... only to find over time that those ribbon cables eventually died and burned SO many of our cards out ( connected to 280x at the time ) ...

now usb risers all the way - 56 of them and growing - and not one hitch - with less power issues and almost NO burnouts of cables or connectors ...

ccminer-spmod runs on all the cards now ( replaced 75% of the farm with gigabyte 750ti oc ( and lp versions ) ) and they run a little more hot than a few months ago ( which im guessing is due to the optimizations ) - but so much more smooth and a higher hashrate on stock clocks ... no oc on any of them - no need really - and way too fiddly with firmware flashing etc with all the cards in the farm ...

if quark is back on with yaamp and has no issues - would it be better to mine there? ... we find that westhash seems to have less downtime than yaamp ...

the test miner ( 2 x gigabyte 750ti oc lp ) which is still donating currently under quark to sp - is STILL rock solid with the -g 5 -i 22 parameters for quark ...

might setup the donation links on various algos on westhash with yaamp being the failover if westhash link dies altogether ( and the link has to die completely for it to failover ) ... or failover to nicehash THEN yaamp as the last failover ... what do you think guys n gals ...

will now pull the latest from git - compile and see how that runs with the same settings ...

see the progress here ... https://www.westhash.com/index.jsp?p=miners&a=12&addr=1CTiNJyoUmbdMRACtteRWXhGqtSETYd6Vd ...

sp - we will keep this miner donating to you for the time being and keep updating to test ... but we need to know which algo you would rather have running on ... if quark is it - then we can leave it where it is for the time being ...

guys n gals - all of your advice would be most helpful ...

tanx ...

#crysx

I noticed slight drop in performance in the new 1.5.48 version on quark. I have 3x750Ti and previous 1.5.47 gives a bit more.

Thanks sp...I looked on my rig ( has 4 black editions ) and they are hashing at a higher rate then the other two GV-N75TOC-2G cards..same oc.
Ps Could you please look into what is causing us not to do round robin bat files with ccminer that keeps on crashing. We need backup pools in the bat file ect. It appears to be in ccminer itself.  
With cudaminer we can also do this a timer but need it for pool backup.

:start
cudaminer.exe -l T5x24 -o stratum+tcp://nyan.mining-pool.co.uk:3336 -u me -p x -d 0,1,2,3 --time-limit 300
goto start

  Thx

Extract the bios from a gforce black. Flash it on the windforce (GV-N75TOC-2G)

Thanks for your support.
For quark you can try -i 22 -g 5 if you have few cards.
On rigs with 6 cards it seems to run slower than with the -g parameter.

I believe yaamp has fixed their issue.  I've been mining for the last couple hours with them (quark -i 22 -g 4) and haven't been seeing extranonce errors (both with a compile before commit a2cc794 and after).

Either can I..but if a stratum goes down then it may happen to me.

Good but you must keep a box fan on them if the temp on your readout app goes above 60c... And the farther away the cards are from each other the better. I'm trying to get a 980 gtx rig and I'm going to space them 30cm apart. When ever I get my first 980 gtx.  
Ps That looks like my rig.  
Sorry sp for messing up your thread.

Yes Take a lazer light temp reading underneath the card and compare it to the temp your evga precision x or what ever you use and you will see a big difference in temperature. Those caps get very very hot.

How do you do that sp?  Thx

bathrobehero Yes these cards need external fans like a box fan or a 9mm fan on each and I don't let my cards ever go above 60c. When they reach this I put a fan on them and when it's hot in the room I set max temperature to 60c. I didn't know this and burned out 2 cards on this computer. The highest temperature for these cards should be 55c to 60c IMO. Another thing is to have the cards no closer than 15 cm  to each other in an open rig. One of those cards still works but won't let me overclock.