Problem with a RIG 6 GPU rx 480 Sapphire

[mina1]

Good afternoon friends,

I have a problem with a rig of 6 GPU, I do not know if someone has happened to something similar.

Rig Details:

6 GPU rx 480 Sapphire

1 moanbord H81 asrock

8 Gb Ram

1 Sdd 64 Gb

1 Power supply RM 1000X

Consumption of the 6 graphs 91 W GPUZ

Windows 10, Claymore's Dual Mining Ethereum + Decred_Siacoin_Lbry AMD + NVIDIA GPU Miner v7.3


My problem is that the GPU 4 (which was changed with the 1-2-3 .... I always miss the GPU4) which is the big slot x16, starts well at 29.2 Mh, but down every moment to stay At 18 MH, after a few hours it is set to 0 MH and the mining program is stopped.

It is always the x16 slot, I do not understand why this slot always fails, and changed the adapter rig, but problem continues.

Attached is a photo of the screenshot.

Greetings, thank you very much.

[mina1]

up

[Raja_MBZ]

Are you using any over-clocking tools? If so, try mining without it.

That seems like the problem of either over-clocking or under-volting.

[twan69666]

It's probably a resources issue. What are your bios settings for that slot? Do you have it on auto, gen1, gen2 or gen3?

[swogerino]

Try resetting that card to stock settings, it appears you have under volt all your cards as they consume 150 watt with stock settings while yours are at 91 watt and the GPU 4 it is not behaving like it should for this under volt. I see other cards are working well, so you just have to bring back to stock factory settings the GPU 4.

[mina1]

Are you using any over-clocking tools? If so, try mining without it.

That seems like the problem of either over-clocking or under-volting.

Good afternoon, thank you very much for your answer first

I do not use any software for overclocking, I modify the room (ATIWinflash) for a cookie whose performance is 29 MH, I use the drivers Radeon Software Crimson Edition 16.7.1. I do not modify any value with any software.

I do not understand why if they all have the same room mood, only fails this if it is subject of over-clocking or under-volting, that this connected to the x16 slot affects something?

I have changed them with others and it always repeats the same problem in the x16 slot, so I discard that it is the subject of the GPU, it has to be something of the motherboard.

Thanks for your time. : D

[mina1]

It's probably a resources issue. What are your bios settings for that slot? Do you have it on auto, gen1, gen2 or gen3?

Good afternoon, thank you very much for your answer,

Do not modify it as it comes from the factory, do you recommend changing it?

Thank you very much: D

[mina1]

Try resetting that card to stock settings, it appears you have under volt all your cards as they consume 150 watt with stock settings while yours are at 91 watt and the GPU 4 it is not behaving like it should for this under volt. I see other cards are working well, so you just have to bring back to stock factory settings the GPU 4.

Good afternoon, thank you very much for your answer,

I do not think it is the subject of the graphics, they are stable and do not give me any error of share, it is something of the x16 slot that does not keep the band wide, because the GPU keeps the W, but the MH.
The only thing I remember is that at the beginning I had 27.5 MH, and since I changed to 29.5 with another room I have experienced this problem.

Thanks for your time, my friend: D

[mina1]

Good again,

Look how weird, if I raise the fan to 100% I recover the memory to 90%, if I let them automatically go down to 60%, does it make any sense with the temperature that is at 75c?

I attach a photo.

A greeting.

[twan69666]

It's probably a resources issue. What are your bios settings for that slot? Do you have it on auto, gen1, gen2 or gen3?

Good afternoon, thank you very much for your answer,

Do not modify it as it comes from the factory, do you recommend changing it?

Thank you very much: D

Since you have 6 cards, have you tried running with 5 cards just to see if the same thing happens? Have you tried removing a card from a x1 slot?

Otherwise, reboot it and go into the BIOS. Change the x16 slot to gen 1. 

[mina1]

Yes, running with 5 GPU no problem, always leaving the x16 slot free.
I'm going to try to change the configuration of the bios to GEN1.
Thank you very much.

[CryptoWatcher420]

Good afternoon friends,

I have a problem with a rig of 6 GPU, I do not know if someone has happened to something similar.

Rig Details:

6 GPU rx 480 Sapphire

1 moanbord H81 asrock

8 Gb Ram

1 Sdd 64 Gb

1 Power supply RM 1000X

Consumption of the 6 graphs 91 W GPUZ

Windows 10, Claymore's Dual Mining Ethereum + Decred_Siacoin_Lbry AMD + NVIDIA GPU Miner v7.3


My problem is that the GPU 4 (which was changed with the 1-2-3 .... I always miss the GPU4) which is the big slot x16, starts well at 29.2 Mh, but down every moment to stay At 18 MH, after a few hours it is set to 0 MH and the mining program is stopped.

It is always the x16 slot, I do not understand why this slot always fails, and changed the adapter rig, but problem continues.

Attached is a photo of the screenshot.

Greetings, thank you very much.

first off gpu-z isn't accurate there is certain things its not taking into consideration for power consumption. Id bet the gpu is consuming more than what you think and what you see. 1000w psu isn't enough especially when you take into consideration of the 80 percent rule, that psu should be at least an 1100 to 1200w psu for 6x rx 480's expecially mining dual mining with claymores

tbh your system is underpowered by atleast 50w which to say the least might be the cause of your issues.

[mina1]

Good afternoon friends,

I have a problem with a rig of 6 GPU, I do not know if someone has happened to something similar.

Rig Details:

6 GPU rx 480 Sapphire

1 moanbord H81 asrock

8 Gb Ram

1 Sdd 64 Gb

1 Power supply RM 1000X

Consumption of the 6 graphs 91 W GPUZ

Windows 10, Claymore's Dual Mining Ethereum + Decred_Siacoin_Lbry AMD + NVIDIA GPU Miner v7.3


My problem is that the GPU 4 (which was changed with the 1-2-3 .... I always miss the GPU4) which is the big slot x16, starts well at 29.2 Mh, but down every moment to stay At 18 MH, after a few hours it is set to 0 MH and the mining program is stopped.

It is always the x16 slot, I do not understand why this slot always fails, and changed the adapter rig, but problem continues.

Attached is a photo of the screenshot.

Greetings, thank you very much.

first off gpu-z isn't accurate there is certain things its not taking into consideration for power consumption. Id bet the gpu is consuming more than what you think and what you see. 1000w psu isn't enough especially when you take into consideration of the 80 percent rule, that psu should be at least an 1100 to 1200w psu for 6x rx 480's expecially mining dual mining with claymores

tbh your system is underpowered by atleast 50w which to say the least might be the cause of your issues.

Good CryptoWatcher420,

I think the source goes to 60% since only ETH, if dual use DCR increases the consumption per GPU to 130 W, then if that would go just for the source, also I confirm that it is the Gold certification 83.3A to + 12V (Corsair RMx Series ™ RM1000x).

Also I can tell you that the fan of the source goes to the minimum.

What I do not understand is that if the GPU works for devajod e 65 C, it works 90%, if it rises to 75 C which is the standard does not exceed 60%.

Thank you very much for the help.

[evakka]

So why don't u try keeping the temperature down for the entire room.
What about the temp of other cards? Above 70 degrees?

Also some other suggestions
- at the card that is plugged to x16 PCI port connect the hdmi or dvi for monitor display
- Are your risers powered? If not u have to get power risers
- if the riser that connects to x16 port is 1x to x16 change it to x16 x16
- try testing other OS than windows to check if this malfunction occurs again

And...
I agree with CryptoWatcher420 for the PSU

[mina1]

So why don't u try keeping the temperature down for the entire room.
What about the temp of other cards? Above 70 degrees?

Also some other suggestions
- at the card that is plugged to x16 PCI port connect the hdmi or dvi for monitor display
- Are your risers powered? If not u have to get power risers
- if the riser that connects to x16 port is 1x to x16 change it to x16 x16
- try testing other OS than windows to check if this malfunction occurs again

And...
I agree with CryptoWatcher420 for the PSU

Good Evakka,

The other GPUs work at about 76 C and are at 29 MH, they have normal behavior.

I have it connected to the HDMI.
The operating system is windows 10, I reinstall it a few days ago to discard that it was SO.

On the power supply, we calculate that if each GPU consumes 90 W but we will go over 110W Real 110W / 12V = 9.1A x 6 (GPU) 54.6 A onli GPU but 150 W Micro Ram ... Total = 67 To the power of this source are a total of 83.3A (there are over 194 W to be the full load of the RM 1000X.

I think my calculations are correct?

I will have to buy an extender x16 to x16 because I do not have them.

Thanks for the reply: D

[xxcsu]

1000wats power supply is not safe for 6 x rx 480 cards, if you dont know how to under volt your cards. Loosk like you overclocked your rx 480...
If gpuz reading is 91 watts , its mean your card can use between 120-150 watts easy ... That 91 watts is only for the gpu power draw , not the memory what you already overclocked to 2080.
Are u using powered risers If yes how many risers is on one 12V cable ?
the molex connector is powered on the motherboard ?

If your cards have a different asic quality , the same tune might not going to work for all of your cards. try to lower the mem frequency for that card , what give you trouble.

[evakka]

So why don't u try keeping the temperature down for the entire room.
What about the temp of other cards? Above 70 degrees?

Also some other suggestions
- at the card that is plugged to x16 PCI port connect the hdmi or dvi for monitor display
- Are your risers powered? If not u have to get power risers
- if the riser that connects to x16 port is 1x to x16 change it to x16 x16
- try testing other OS than windows to check if this malfunction occurs again

And...
I agree with CryptoWatcher420 for the PSU

Good Evakka,

The other GPUs work at about 76 C and are at 29 MH, they have normal behavior.

I have it connected to the HDMI.
The operating system is windows 10, I reinstall it a few days ago to discard that it was SO.

On the power supply, we calculate that if each GPU consumes 90 W but we will go over 110W Real 110W / 12V = 9.1A x 6 (GPU) 54.6 A onli GPU but 150 W Micro Ram ... Total = 67 To the power of this source are a total of 83.3A (there are over 194 W to be the full load of the RM 1000X.

I think my calculations are correct?

I will have to buy an extender x16 to x16 because I do not have them.

Thanks for the reply: D

Hi again Mina,
Regarding the operating system you should try other than Windows, and i meant Linux.
Do a Search in the forum for simplemining.net, it is very simple to install it and it needs just a USB stick.

Try it and check the faulty GPU

[mina1]

So why don't u try keeping the temperature down for the entire room.
What about the temp of other cards? Above 70 degrees?

Also some other suggestions
- at the card that is plugged to x16 PCI port connect the hdmi or dvi for monitor display
- Are your risers powered? If not u have to get power risers
- if the riser that connects to x16 port is 1x to x16 change it to x16 x16
- try testing other OS than windows to check if this malfunction occurs again

And...
I agree with CryptoWatcher420 for the PSU

Good Evakka,

The other GPUs work at about 76 C and are at 29 MH, they have normal behavior.

I have it connected to the HDMI.
The operating system is windows 10, I reinstall it a few days ago to discard that it was SO.

On the power supply, we calculate that if each GPU consumes 90 W but we will go over 110W Real 110W / 12V = 9.1A x 6 (GPU) 54.6 A onli GPU but 150 W Micro Ram ... Total = 67 To the power of this source are a total of 83.3A (there are over 194 W to be the full load of the RM 1000X.

I think my calculations are correct?

I will have to buy an extender x16 to x16 because I do not have them.

Thanks for the reply: D

Hi again Mina,
Regarding the operating system you should try other than Windows, and i meant Linux.
Do a Search in the forum for simplemining.net, it is very simple to install it and it needs just a USB stick.

Try it and check the faulty GPU

Thanks for your time, I'll perfect prove SimpleMining.net.

Thank you very much.


 

[koody]

Use pci-e USB adapter, maybe your slot 16x is bad.

[mina1]

1000wats power supply is not safe for 6 x rx 480 cards, if you dont know how to under volt your cards. Loosk like you overclocked your rx 480...
If gpuz reading is 91 watts , its mean your card can use between 120-150 watts easy ... That 91 watts is only for the gpu power draw , not the memory what you already overclocked to 2080.
Are u using powered risers If yes how many risers is on one 12V cable ?
the molex connector is powered on the motherboard ?

If your cards have a different asic quality , the same tune might not going to work for all of your cards. try to lower the mem frequency for that card , what give you trouble.

Good Xxcsu,

I understand, well try to lower the frequency of memory.

If I use 6 risers.

The molex connector on the motherboard I do not use it, I do not have it connected (neither physically nor by the BIOS) But I have always been left with the question of whether it is necessary or not?

Thanks for your answer: D