6600 XT Strange Issue

[fitzuk]

Hey guys,

I am having a strange issue with the 6600 XT cards.

I have built 2 rigs now with them with no issue, hive picks them up and everything is fine.

I am trying to help a friend build the same rig with the exact same components and I can’t get hive to recognise the cards, occasionally it will pick a card up. But when rebooted it doesn’t or when an extra card is added it essentially kills all the cards. I have tried the latest stable image as well as the latest beta, tried updating to the latest in the gui etc.

The card does pick up fine in the 16x port which would indicate an issue either with the risers or the bios but I have confirmed the risers & mother board working with another rig. I just can’t figure out what the issue is.

Motherboard: H110 Pro BTC (tried 3 different ones)
Risers: Confirmed working with other cards / rigs

They show as “Navi 23 [Radeon RX 6600/6600 XT/6600M]” and spit out a ton of errors when booting up. They do occasionally start mining with default settings.

[martlabre]

You need the latest Hivos #140

[fitzuk]

Hi,

Thanks for the reply.

Unfortunately I have tried that version and pretty much every other possible combination.

I have another rig that is currently working with the 6600 XT with 13 cards. All the same components.

I have taken this hard drive and attached it to the known working rig and it worked fine (ruling out a HiveOS issue)

But my issue is I now have absolutely no idea why it doesn’t work. I have changed risers with fresh batches of risers and ones from known working rigs, I have swapped out the PSU, I have changed the pico & the cables.

It only leaves the cards. What’s interesting is i have got it working with 2 cards and they appear correct in HiveOS typically if I go to the bios first and then exit straight into the OS they appear but then I can restart and then they go back to not working or I can add an a extra card and it also kills the other 2 and then all 3 show as “Navi 23”. The risers are all firmly seated.

Could this be a motherboard setting I have missed or anything else I haven’t thought of?

Thanks in advance

[ct21dnn]

PLS check H110 pro+ motherboard bios setting.
You must connect two power sockets (Molex power sockets on the motherboard). Otherwise, there will not be enough power to run all 13 PCIE slots.

First, reset the BIOS by removing the BIOS battery (wait at least 30 seconds)
Connect only one GPU to the x16 slot (the longest one). Power on (button on the motherboard) and enter the BIOS. Modify the value in the BIOS as follows:
OC Tweaker\Voltage Conf:
PCH VoH: 1200 mV
VCCIO: 1200 mV

Advanced\Chipset Conf:
Above 4GB mmio: enabled
VT-d: disabled
All PCIE link speeds: Gen2
All ASPM support: disabled
IOAPIC 24-119: Enable
IGPU: enabled
HD audio: disabled
Reset power consumption: turn on

-Save and exit the BIOS

-Use only one GPU to run your device

If everything is ok (1 GPU is mining), now you can pass the remaining 12 GPUs to the stage
-Restart and enter the BIOS. Disable CSM in the boot section.
-Save and exit. After restarting, power off before starting! (This is important)

-Now connect all GPUs to the remaining x1 slots (short slots). Check that all GPU??? risers and power connections are provided and correct.

-Turn on suddenly again. All 13 GPUs should be recognized and work normally.
I hope you can solve the problem.

[aharrison]

I had the same issue a few days ago. Check your power cables. For me it was a faulty power cable powering the riser, specifically one end of a splitter, the other end was fine. I troubleshooted for days before I tried the power cable.

Make sure you are providing power and enough of it to both the riser and the GPU. Also, Are you supplying power to the risers with SATA? They only provide 55 W which could be an issue as well.

[Cekickojisvira]

I had the same problem on etherminer. I solved the problem by installing version 0.6-210-beta210914, then upgrading to 0.6-210@211020 and switching to nbminer. Now everything is working properly