Quote from: lesjokolat on June 16, 2018, 08:04:46 PM

Quote from: 4play on June 16, 2018, 07:51:16 PM

Quote from: Teti2347 on June 16, 2018, 07:42:23 PM

Hey Im hoping someone can help me out with an "out of memory" error I keep getting. I have a rig with the following build:
8 X NVIDIA 1070s
Asus H270-P motherboard
2X EVGA 750 PSUs
Windows 10
120 GB harddrive
8GB RAM

I am mining with PoenixMiner and using Nanopool and keep getting an error message. (I have 2 other rigs built the exact same and havent had any issues with them). Yesterday it was mining fine and the the program launched out, so I reset the rig and have been having issues since. I was able to get 7 GPUs running before but when I plugged in the 8th card, its giving me an error code. I tried switching out the risers and now am only able to get 6 cards working before the same error occurs -- "CUDA error in CudaProgram.cu:377 : out of memory (2)"

MSI Settings:
Power: 64
Temp: 67
Core Clock: 0
Memory Clock: 585

Any ideas??

I have also 1070 gpus and i using 2x 850W PSU.

My MSI OC Settings are:
Power 65
Temp 76
Core Clock 0
Memory Clock 710 -> The only exception is one gpu because I had problems with this CUDA error. This gpu have memory clock 650 other settings are same.

https://imgur.com/NravSg9.png

try increasing your swap memory to at least 24 gig

]

Was having same issue right after the Apr 2018 update. Current rig is 8 1060s with 1000W PSU. Change swap file to 22GB (22000 MB) and 3.0c started working as expected.

Thanks for the tip...

oops... PM died after about 10 min.. Up'd the swap to 25 GB.