Rig Random GPU Not responding

[fisch123]

Hi,

I am using 9x 3060TI. After a few min or hours one GPU Starts not to responding.

OC settingen are
-500 core
2000 hz
130W

Normal Settings which should work. All GPU and Riser are connected by PCI cables.

I am Using a 1500W and 1000W PSU on an Asus Z170 Gaming. It Runs with a SSD

Sometimes it doesnt even Show an Error Message, the Rig Just restarts. But by restarting it Freezes sometimes. When i Check the Log (which ist created every 2nd or 3rd time) it says “wdog Threads(s) Not responding”

Can you Help?

[nanial]

Make sure you balance the wattage between PSU’s and second put your core at -200 and check if you have the same problem…

[fisch123]

You mean Not all GPU on one PSU? I have on the 1500W Platimax 5x 3060Ti +SSD+ Mainboard (Pentium 4560G) and on the 1000W DARK Power 4x 3060TI + 8 fans

[fisch123]

-200 clock doesnt help

[adilboztas]

Most likely the problem is with the PSU.

[nanial]

Check all the cables in all Psu’s (to the cards and to risers) … make sure u don’t have burning staff… In your situation a make all the connections from the beggining to make sure is all ok… this error (freeze) is hardware error… (maybe one card have a problem…, maybe your memory (reseat change slot ) have problem or your ssd …) search !!! Freezing your rig without error
is Hardware Problem…

[fisch123]

It’s just freezing by restarting…and it Just restarts because of losing a GPU.

OK i’ll try to Test every part and give you a feedback

[fisch123]

I have checked all cables again. And I really can`t find the failure. Mostly GPU 3 is lost. I tried it with different PSU now, GPU 3 is connected by a ricer to a PCI 3.0 1x

Also when I use a Kolink adapter to split the (2nd slot) 1x PCIE x16 to 4x PCIE the GPU 3 is mostly lost…what I want to say with that, it cant be the 3rd PCIE slot because GPU 3 is connected to the 2nd PCIE slot by the Kolink adapter (just a few times it is another GPU). Furthermore the rig doesnt start as soon I connect 4 GPU to the Kolink Adapter (i have also testet it with a 2nd adapter - same problem).

The error log does mostly show me this:

GPU 3 still has 133W Power, but its hashrate is 0... (also when i change the GPU, its always GPU 3)

sometimes it `s another GPU …