Ethereum Mining Unstable Above ~150Mh/s Regardless of Hardware Setup.

[deeee]

Howdy. I'

I have an ASRock H110 Pro BTC+ and 9 AMD GPU's (7 RX 480's and 2 RX 580's). For awhile, I found I could only get the machine to run consistently with 6 of the GPU's plugged in. I assumed that something was wrong with the remaining 3.

However, after some experimentation, I've found swapping cards, risers, etc seems to have no effect on stability. Moreover, if I lower mining intensity, but add more cards back in, I get about the same number of Mh/s.

I seem to hit a ceiling somewhere between 150 and 160 Mh/s. Anything more than that and my machine will randomly reboot after a few seconds of mining. Sometimes it reboots as soon as all the DAG's are loaded, sometimes a minute later, but it never stays up long.

I currently have the cards all overclocked, but I got the same results when I briefly flashed them back to factory. Nothing seems to keep them online.

I've tried both ethminer and Claymore -mode 1. I'm running AMD's 17.50 Linux drivers under Ubuntu 17.4.

Is there some piece to the puzzle that I'm overlooking?

[Marvell2]

random reboots are either not enough power or oc is too high on some of your cards

[deeee]

Thanks for the quick reply!

I've got a 1500W PSU  plugged in (Corsair AX1500i). Monitoring the power usage doesn't show any usage anywhere near that limit.

As for overclocking, I thought about this, but as I mentioned, even running stock I have this problem.

I can run 6 GPU's each around 27Mh/s no problem (just over 150Mh/s total). I can also run 7 GPUs at around 23Mh/s without any real issue (same average total).