Hi all. I have 8 AMD 6800 cards in my case. I recently changed the algorithm on the teamred miner and it started giving an error for one card.
Memory access fault by GPU node-4 (Agent handle: 0x555557138e90) on address 0x7f4fbd2c4000. Reason: Page not present or supervisor privilege.
Please help, has anyone encountered something similar.
Try setting more conservative clocks/voltages after each crash and reboot. Repeat until stable
I am actually experiencing the same thing with my 6600 rig. I am not able to mine any other coin now because of this problem. I’ve tried all manner of overclock/algo combinations, and nothing works. They won’t even mine with absolutely no OC’s. The only coin they are able to mine now is Meowcoin. Which don’t get me wrong, I like the coin, but I would like to be able to mine something else. Hopefully someone has run into this issue and has a fix for it.
No oc is typically the same as bad/poor oc’s. Try lowering oc to very conservative settings, such as ~1000 core clock ~800mv core voltage and 800mhz memory clock, reboot and see what happens
Thank you, still working on this. One of your Tech support people led me to realize I didn’t have enough Ram in my Octominer (the guy I bought it from ganked a 4g stick). So now I’ve gotten several of the cards up and mining Nexa, but still have a few picky cards that don’t seem to want to mine it. I’ll respond after trying those settings you mentioned.
One card still doesn’t want to mine Nexa with those conservative settings. (Again, it mines Meowcoin fine). What’s even more of a conservative setting to use? I just don’t know how low or high I can go with the settings.
Those settings should be fine. Are those cards not mining still having the errors that you were getting earlier?
