Quote from: f00ch0w on July 02, 2017, 12:40:05 PM

Hey there, having an issue with most of my rigs with what I'm struggling for days now, wanted to give it a shot and post here to see if someone has an answer.

Got 5 rigs with this setup:

MB: ASROCK H81 PRO BTC rev 2.0
CPU: Intel Celeron
PSU: Antec 750W Gold
GPUs: 6x Gigabyte GTX 1050ti 4GB
Risers: 6x 16x USB Molex powered risers.
OS: nvOC v0017
Clocks: 150/600
Coin: ZEC

Few rigs with:
MB: ASROCK H81 PRO BTC rev 2.0
CPU: Intel Celeron
PSU: Antec 750W Gold
GPUs: 3x Gigabyte GTX 1050ti 4GB and 3x Gigabyte GTX 1060 6GB
Risers: 6x 16x USB Molex powered risers.
OS: nvOC v0017
Clocks: 150/600
Coin: ZEC

Rigs crashing all the time, most crash with GPU1 stopped working, then they crash one by one. Sometimes they exit with thread errors.

What I've tried:

1. Lowering the clocks
2. Using the default clocks
3. Using Powerlimit (Both lowering and increasing from it's default value by 5-15W)
4. Powering the PCI buses with 2x Molex connectors to the both (With and without it)
5. Swapping the risers on GPU that crashed first
6. Swapping the GPU that crashed first
7. Swapping the PSU
8. Powering 2x GPUs per PSU cable
9. Using Auto/Gen1/Gen2 for PCI bandwidth speed

Currently, Im out of ideas what to do. Anyone has a suggestion?

Try running one graphics card in the full slot (if they have a full slot) and run it for, see it lasts half an hour, then add a second one.. preferably in a full slot (before trying screw around with risers and stuff) and build up with half an hour intervals seeing at what point cards become unstable.. It may be PSU isnt able to supply the power and when it hits its thermal switch it brings down the output.. killing your cards.. locking up the system..