Bitcoin Forum

hi guy, I am using nicehash legacy.

My rig always restart if I set my afterburner power limit to 90, now I am running 80. Got fews times after I install the Gigabypte PSU (power limit 90), I saw 3 of the GPU stop working after few seconds

Also I am using NH to mine, so if I turn on Lyra2 my rig also will hang and restart it self, now I only can only Equihash and Neoscrypt.

May i know what is the problem?

My rig spec:
-Asus EX-B250-V7 x 1
-Intel G4400, 3.3 GHZ , 3MB Cache, LGA1151 ( BX80662G4400)
-Corsair ForceGT 120GB SSD x 1
-Corsair HX1200 x 1
-Gigabyte Xtreme Gaming 1200w x 1
-MSI GeForce GTX 1080ti x 6
-Kingston DDR4 2133MHz 4GB x 1

I connect Corsair HX1200 to motherboard,SSD, CPU, 3 riser and 3 GPU, and Gigabyte Xtreme Gaming 1200w to 3 riser and 3 GPU.

MSI Afterburner Setting For 2 PSU IF I set the power limit to 90, rig will restart it self or power on the 2nd PSU will trip/cut off.
Power Limit - 80
Temp. Limit - 80
Core Clock - +95
Memory Clock - +600
Fan Speed - 80

MSI Afterburner Setting For 1 PSU( work fine)
Power Limit - 100
Temp. Limit - 80
Core Clock - +95
Memory Clock - +600
Fan Speed - 80

Anyone have this problem?Huh

**Update
Today reformat my rig and add new ram to 8 gb, and new SSD. And the problem still the same

Problem Solved
For restart - change riser and SATA cable port
For trip - PSU failure

Could be a problem with afterburner that's causing an error i had similar problems with windows switched to ethos and never had anymore problems.

I try to format the whole C: , and the problem is still here.... even change to 8gb ram and new SSD.

Set virtual memory to at least 16Gb.

Emm.. Why you need 2x 1200W psu for powering 6 GPU's? ???

he set them to 100% power

that is 6 x 250 = 1500 watts


@ op  set them to 80%

You cannot find the problem if you have to much variables in your equation.

First you must simplify your rig, throw out second PSU and leave only 3 GPU and HX1200.
-At this point if you rig is restarting it could be unstable voltage or memory problem.

After you see that all is working without restarts add another PSU and only one card, after one day, add another card, after one more day add last card.

OMG what are ya doing here?

because 1xMSI1080ti need around 250watt

Cut power usage then. This cards consuming 250 wt will never recoup at current prices.

Are you using the cable to connect both PSU btw them?

Have you tried to switch PSUs position?

-Gigabyte Xtreme Gaming 1200w to a motherboard, SSD, CPU, 3 risers and 3 GPU

-Corsair HX1200 to 3 riser and 3 GPU.

How are you connecting that two PSU? Using add2 PSU (or something similar)? Or jump-start?

I would agree. The problem looks like leakage current between PSUs. You should separate power lines from different PSUs by using add2PSU with relay. Also GPU's riser power should be connected to the same PSU which the GPU connected

Actually not. All risers and MB should be connected to the same PSU.
Better  to use relay starter for second PSU.
Better have the common ground for both PSU.

So using 2 or 3 PSU is not such easy, better use this only if urgently needed

Totally he is using the 2400 watts of SMPS to get the power supply but I do not find the people having issue with this much power supply. Actually what is the card he is using now and confirm the clocking he has done here.
If all perfect we may suggest him to go with the Windows OS to Linux.
Even there are some OS perfectly meant for mining sole purpose only mate. Please cable connected the riser to board as well once.

Its always problems with multiple PSU`s it seems. Shouldn`t you connect all risers + motherboard on same PSU?

Yep  it's necessary for obviously reason.

Yes, I am using cable to connect them.

Like the link below:
https://www.ebay.com/itm/Dual-PSU-Cable-Power-Supply-24Pin-ATX-Motherboard-Mainboard-Adapter-Connector-2/232665489120?epid=7014841479&hash=item362bf19ee0:g:3bUAAOSwjuNahF27

I've been running risers/mobo on an 800W PSU and all GPU power on a separate 1200W server PSU on 4 different 8 GPU rigs for a year and a half now. Never had any issues with that setup. Switch on GPU power first, then mobo. No add2psu or crap needed. People warning about this setup have never tried to run it.

You may  burned some PCI-e port  on MB if you did not feed all your risers and MB from the separate PSU for sure.

What do you mean dimpsk? is it you mean all risers must power with the same PSU as MB?

Yep, it is necessary. How ya 'd balance GPU supplie does not matter.

Basically we have 2 groups of people


Group 1 - Power MBO, HDD + all riser from one PSU and all the cards from other PSU
Group 2 - Power MBO , HDD from one PSU and all the cards with their riser from other PSU

I Don't know which one is correct, because all options will work until you don't get difference in electric current then it will all go to smoke.

Both work indeed, yes I do both of that scheme. It depends on conditions when we have PSUs with big enough wattage, powering the risers and GPUs from the same source is better. It is about power management, optimizing and risk management.

Make that a Group 3. The best way is to use one PSU for everything. No smokin :-)

There is no  harm in first case at all.
But. Ya can load first PSU with some GPU in addition,
The point was: power supply of MB and all risers should came from the same PSU

That's it