Problems with high GPU count

[Amstellodamois]

One of my PSUs was DOA, so I've been running 9 GPUs with success until today. I received a replacement PSU and connected the last three GPUs today.

...and that's one hell of a shit storm.

I finally got the 12 1080 Ti to be recognized in the device manager but many miners aren't so happy with all of these. I wonder if MSI Afterburner can handle that many cards.

https://image.ibb.co/h0XWNc/2.png


Any advice?
(Other than "just split your cards into two rigs")

[1714952116]

1714952116

[1714952116]

1714952116

[1714952116]

1714952116

[Amstellodamois]

Sometimes I'm happy: https://image.ibb.co/hetT3H/Screenshot_20180225_011558.png

Most often, I'm not : https://image.ibb.co/jhaaiH/Screenshot_20180224_205918.png

[Bare]

Sorry for being "that" guy, but serves you right for keeping your cards temps at 72-76°C...

If more people kept their cards at least below 70°C for 24/7 mining, the world would have way less crash/freeze and issues with mining...etc. threads

I'm considering that a prerequisite before troubleshooting anything!

either make sure your cards have proper ventilation or tune down your cards a bit, but in ANYWAY get them below 70°C for 24/7 mining and don't let anybody tell you otherwise!
In my opinion, 63-64°C is a sweet spot, of course lower is better if you can.

[Wipro]

Please do not overclock the card too much to get the high profit. It will reduce your cards life span and as above person said you have to keep your card at the maximum of 65 to 70% degree heat.

This heat issue will be resolved if you dedicate the 200 watt for each card. For example, if you have 1000 watt SMPS. Please add upto 5 cards to avoid the heat issue. Sometimes power cables gonna melt too. So check it on weekly basis.

Tip: Turn off your rig for an hour of time every week. This is good for your cards and mobo.

[Amstellodamois]

Guys, I don't have a heat problem.

It's not that the mining stops after a while, it just won't start.

[philipma1957]

One of my PSUs was DOA, so I've been running 9 GPUs with success until today. I received a replacement PSU and connected the last three GPUs today.

...and that's one hell of a shit storm.

I finally got the 12 1080 Ti to be recognized in the device manager but many miners aren't so happy with all of these. I wonder if MSI Afterburner can handle that many cards.




Any advice?
(Other than "just split your cards into two rigs")

yes  go to the monitor section of msi afterburner.

it has a stock setting of 1000 micro seconds  so it samples all 12 cards 60 times in a minute.  each card has  at least:

12  things to look at  so 12 x 12 x 60 = 8640  things are checked every minute = a shit ton of lag time    unless  you have an i7  cpu.


so set the  sample time to 12000 micro seconds    that is   only 12 x 12 x 5 =  720 things every minute


this will help.  if you need a photo of the screen in question    I can provide it.

[Amstellodamois]

Thanks, I know that screen.

I had changed to 2000 ms earlier, I switched to 5 seconds just now. I'm not seeing more cards appear in the monitoring section, I was hoping it would. Afterburner sees all the cards, though: they all appear in the info section.



About my stability issue, I might have isolated the problem: the rig has been mining all afternoon with 11 of my 12 cards, so I think that last riser (or card ) is bad. Fortunately, I have ten spares

[Bimmber]

I wonder if MSI Afterburner can handle that many cards.

As far as I know, AB does not support more than 8 cards. Use nvidia inspector, I prefer it much better.