i don't know if this is available under windows but in the nvidia suite there is something called nvidia-smi, it gives a detailed overview of what's happening with your cards and you can see how much they use (in terms of power), how much they are used by the system, temps, fans, ... With this you can see if one of the card is not responding and try to swap cables, cards, ... If not, then the usual debugging of mining for a couple of hours with card, then another one, ...until it starts crashing again.

it is profitable if you can get a decent price for GPUs and have low power costs. I started 8 months ago while ETH was still around 200$, it was already profitable at that time but return on investment for the hardware was longer. Now it's way more profitable than 8 months ago.

Has someone managed to run more than 13 GPU (nvidia) on Linux ? I currently have an Asrock card with 13 GPU from Nvidia and was tempted to go for a 19 gpu mining rig but before doing this, I'd like to make sure it works. I wrote to Asus and the only thing they can provide is a recommendation but with their recommendation you never reach 19 GPU of the same brand so it defeats the purpose of going for this card in my case.

mines are vertical too but I have placed them on the outside of the rig so that the air sucked from fans is not taking warm air from the backplate of another gpu, This way I keep temps between 45 and 55°C.

bigdigtaker: that's exactly the setup I have

honestly if you plan to mine with only 2 to 4 gpu it's probably not worth it, you'll have to pay for all the hardware and having only a few gpu to get back your money. Of course if you plan to have GPUs like GTX 1080, that's a different story.

It's perfectly fine, I have 13 of them running in the same rig and I get good results with low power consumption

They are probably right not to do anything about it, I'm not sure mining as we know it today will still exist in one or two years. The model is not viable, wasting so much electricity is not something sustainable on the longer run. Like you wrote earlier, there will be one day where the second hand market will be flooded with gpus used for mining because nobody will want those cards anymore, price will strongly go down and only those willing to buy second hand mining cards will be happy. Sellers and manufacturers will both lose something in that game.

They probably don't believe in mining on the longer run. Designing new cards takes months if not years, creating a specific product line for mining is an investment for them and they don't seem to be in a hurry to enter that market. There probably also a question of production capacity, mining is driven by euphoria and you can't match your production capacity to euphoria.

@kostickhole92 : have you managed to make this work?

No I have 13GPU so I don't know what's happening above 13. Are you using 13 GPU from nvidia?

do you have the same result when you try them one by one ? Have you tried other pcie slots?

Oh really? Would a splitter work here?

Thanks, I was going to go for option B but the thread doesn't mention anything about consequences or restrictions linked to this type of connector. I have powered risers already.

Hello,

I have a Asrock H110 BTC+ motherboard with a M.2 SSD connector. I've seen risers allowing to plug a GPU on that connector but before buying, I'd like to know if there are any restrictions or things I should be aware of.
My motherboard has 13 PCIe slots but one of them is making my rig unstable so instead of getting rid of my GPU, I'd like to use it through the M.2 connector.

Thanks
Laurent

Has anyone tried the M2 slot to insert a riser there? If it works that could bring the amount of cards to 14 but are there any restrictions?

I have 6 pins connected to the GPU and 6 pins connected to the riser, would you rather advise to have risers with molex connectors ? Is the 6 + 6 setup less stable?

same kind of issue. I have tried to raise power limit up to a certain limit but it didn't solve the issue. I should probably try to remove power limit but then the whole rig consumes way too much electricity and becomes too hot (more power = more heat).

I have this type of motherboard with 13 GPU running under Linux but I must have a faulty motherboard because some slot are not giving the best performance. If I swap cards or risers, it doesn't change anything, some slots are not performing very well. If I take a card and riser that is performing well on one slot and move it to a slow performing slot, my hash rate goes down while it was higher with the same card + riser on another slot.

I also have one of my slot that makes the whole rig unstable for some reason. If I use that slot, the rig crashes after a couple of hours while when I'm not using it I can mine a week without a crash.

I contacted Asrock and they suggest to return the motherboard. Have any of you had a similar experience?

bought 3 Zotac 3Gb last week, all Samsung.

Well, I have 13 GPU on  a 13 GPU motherboard so it's now like I have much choice. I have an Asrock motherboard, not an Asus one. I'm running the rig with Ubuntu 16 and it's a 100% nvidia rig (see my signature)