Bitcoin Forum

Just built my first rig.
 
Kingston SSD 120GB
Biostar TB85 1150 mobo
6x RX480 Sapphire Nitro 8GB graphics cards
Deep in the mines frame with usb risers
EVGA 1300W Gold PSU

So here is the deal. I had a legit version of Windows 7, upgraded for free to windows 10. Started from there with installation of drivers. No matter what I have tried, if I plug one of the risers into the big PCIEx16 slot, the system will not load, no POST message, no signal sent to the screen. If I plug in a card to the PCIEx16 slot while windows is on, it crashes.

I have tried switching to gen 1 and gen 2 on the mobo BIOS. Switching a card I know works to this PCIEx16 slot. Just running the PCIEx16 slot and not the smaller ones. Tried a few versions of AMD drivers.

For what its worth, all 5 of the cards will have their fans stop right before windows loads to the desktop, then they kick back on. Almost like the system won't load until the fans stop. Claymore dual is working fine with Decred and Eth, but I don't dare mess with OC'ing these things until the 6th card works.

I am powering the rx480's with cords direct from the power supply unit. The risers are being powered by cables that I also attached to the PSU. The only thing I could think is that I have only 1 AUX power cord plugged into the board, the second doesn't fit next to one of the PCI slots.

Please let me know hat else I can try to use my 6th card.

EDIT: Actually randomly after a reboot, only 4 are found now and the 4th is hashing at a measly 4 m/hs. Ugh

Post your rig images

Did you check the PCI latency?  Needs to be set at 96.

Check this out, might help with the BIOS settings:

https://blockoperations.com/configuring-bios-settings-gpu-mining-intel-motherboards/


I just built my rig about a month ago with that board and it's given me all sorts of headaches.  I've had the system lock up and the BIOS settings reset.  Reset CMOS, replaced the battery, and still happens on occasion on hard reboots.  When the system shuts down properly it's no problem, but when one or more cards become unresponsive, it can hard lock up.  I'm using Claymore's miner and uses the flag to reboot when there's issues with a GPU but it will still lock up on occasion.

Thank you for your replies. What I did was installed this older driver: 16.11.3-nov8. Then I disabled automatic driver updates on windows and through AMD. Followed the BIOS settings posted by z92747392. And then the final trick was this. I tried from the beginning with just my one card plugged into the PCIex16 slot. Booted up and noticed it hung at Windows load as usual but I decided to let it do its thing. After the screen turned on and off like 6 times, it stopped and it was stable and this was the first time I had a card be read from this slot. I then restarted with 2 cards, then 3 and onwards until all six. Then I was getting a BSOD.

I uninstalled the AMD drivers and rebooted. I plugged in the last card just to see, and it was recognized by Windows. I then used the AMD cleaner utility and then installed 16.9.2 from guru3d. Unbelievably it now works. Getting about 24m/hs on ethereum and 4400 on decred.

So I guess it boils down to installation methodology. For this board, that driver works. You should install windows 10 with all 6 cards attached and use that driver. Would have saved me a day, but oh well.

On to overclocking! Please share if anyone has a good writeup. The one I found was for the nitro + which I do not have. I have the nitro 8gb non plus (samsung though).