nvidia 1070-1080 mix ring crash

[carlo_0000]

hello
so i have an 1070+1080 ring   but got a lot of trouble
computer freeze or reboot

i keep an eye and notice this

CcminerPalgin, and other version of ccminer tested too

[2017-12-11 15:24:36] GPU #3: GeForce GTX 1080, 667.09 MH/s
[2017-12-11 15:24:36] GPU #3: an illegal memory access was encountered
[2017-12-11 15:24:37] GPU #1: GeForce GTX 1070, 542.26 MH/s
[2017-12-11 15:24:37] GPU #1: an illegal memory access was encountered
[2017-12-11 15:24:37] GPU #4: an illegal memory access was encountered
[2017-12-11 15:24:37] GPU #0: an illegal memory access was encountered
[2017-12-11 15:24:37] GPU #2: an illegal memory access was encountered
[2017-12-11 15:24:38] GPU #5: GeForce GTX 1080, 435.39 MH/s
[2017-12-11 15:24:38] GPU #5: an illegal memory access was encountered
[2017-12-11 15:24:38] GPU #0: an illegal memory access was encountered
[2017-12-11 15:24:38] GPU #3: an illegal memory access was encountered
[2017-12-11 15:24:38] GPU #1: an illegal memory access was encountered
[2017-12-11 15:24:39] GPU #3: an illegal memory access was encountered
[2017-12-11 15:24:39] GPU #2: an illegal memory access was encountered
[2017-12-11 15:24:39] GPU #0: an illegal memory access was encountered
[2017-12-11 15:24:39] GPU #4: an illegal memory access was encountered
[2017-12-11 15:24:39] GPU #1: an illegal memory access was encountered
[2017-12-11 15:24:40] GPU #5: an illegal memory access was encountered
[2017-12-11 15:24:40] GPU #2: GeForce GTX 1070, 769.66 MH/s

look like all gpu's got memory error, they currently at stock speed
using last version of win10 and driver 388.43

i m now running it on eth

[1714191244]

1714191244

[1714191244]

1714191244

[1714191244]

1714191244

[1714191244]

1714191244

[1714191244]

1714191244

[1714191244]

1714191244

[X299]

Change the miner.

[Vann]

Check your temperatures, especially on the 1080's which run hot. I get crashes and reboots when the temp exceeds the temp limit. Reducing the power limit and setting the fan speed using a custom profile is a good way to lower the temperature.

[carlo_0000]

it s not the temp  it s very cold here, and i already use msi afterburner to get higher fan's

i tested eth it work fine

CcminerKlaust             Groestl  work also fine for now

edit

there is definitely something  not right
using multipool miner  i notice CcminerKlaust close itseld (so i guess it crash or so?)

than it started claymore eth dual, and i notice 2 gpu's are gone, they still show up in device manager, i can disable and reenable them
i also already checked the power and risers

CUDA initializing...

NVIDIA Cards available: 4
CUDA Driver Version/Runtime Version: 9.1/8.0
GPU #0: GeForce GTX 1070, 8192 MB available, 15 compute units, capability: 6.1

GPU #1: GeForce GTX 1080, 8192 MB available, 20 compute units, capability: 6.1

GPU #2: GeForce GTX 1070, 8192 MB available, 15 compute units, capability: 6.1

GPU #3: GeForce GTX 1080, 8192 MB available, 20 compute units, capability: 6.1

Total cards: 4
NVML API: failed to get device handle 3, error 15.
NVML API: failed to get device handle 4, error 15.
NVML API: failed to get device handle 3, error 15.
NVML API: failed to get device handle 4, error 15.
NVML API: failed to get device handle 3, error 15.
NVML API: failed to get device handle 4, error 15.
NVML API: failed to get device handle 3, error 15.
NVML API: failed to get device handle 4, error 15.

[carlo_0000]

no idea how much the intensity is
will take a look at that later

maybe motherboard faulty as it is a 5 pice slot i add an 1xto4 pcis slot  to put 8 gpus but motherboard won't boot with more than 6

so i m planing to remove 1 and separate the 1070 and 1080 on 2 different ring, the motherboard is gigabyte 970a ds3p


edit:

i found the problem
i forgot to increase my swapfile  after adding more gpu's