It sounds like the cards are not functioning properly, like I said before. If you tried plugging them into multiple systems that are able to run identical cards in the same slots, but these cards are unable to do anything then that would indicate that the cards are in need of repair. You can try running them on a new OS, but since you are in windows 10 are you able to see the GPUs under device manager? Since they are receiving power, and the fans are turning it is possible that the drivers are just not installed properly, but device manager should help us get a clearer picture of your situation.

Even turning off the PSU is not enough to prevent shocking the cards, even shocks that you are unable to physically feel, because of your threshold for "pain", are well above the amount necessary to kill computer components. Anti-Static bands, discharging your static electricity and making sure you have clean hands when you get inside your computer are a must. Were these cards working properly just a few days ago or have they always been giving you a tough time? What happened between when they were working and when they no longer were functioning as intended?

Have you attempted to connect the GPUs one at a time to make sure they individually are functional? Are you able to place the 2 GPUs that are not functioning properly into another system to see if it is able to get them working? What OS are you running, and does the OS detect the GPUs at all?

It sounds like it would be the cards, maybe they died during a bad flash recently or got a good shock while you were placing them in without discharging your static electricity. If they are unable to work in a slot that other cards are working in, then it probably isn't the slot or the motherboard. The PSU should have enough juice.

Here you can see most of cards benchmarked on cryptonight algo (data may be outdated after the fork - but shouldnt have huge difference)

http://monerobenchmarks.info/

From my experience, there shouldn't really be any kind of extra ordinary difference b/w gens of graphics card as far as mining is concerned. You can check older cards and see for yourself, one major aspect is the power draw - which sometimes can get lower or higher (depending on specs), but I dont think it is much of an issue when considering the total performance h/s/w wise, also taking into consideration current price/difficutly etc..

If you are getting into mining game, make sure you are pretty sure and confident that it will give u the calculated ROI.

Once again only NVidia miners and no AMD?
That's absolutely no good...

Hi guys,

Any one out there was successful for GPU mining under Linux with Tesla K80 cards? I installed NVIDIA drivers and CUDA toolkit and compiled ccminer from source but when I try to launch the miner on any algorithm I always get error messages like;

for neoscrypt algo: Cuda error in func 'neoscrypt_init' at line 1482 : invalid device symbol.

Some others algos: no kernel image is available for execution on the device

I've downloaded the latest drivers and the latest CUDA Toolkit version (9.1).

I tested it under Tesla M60 and all worked well, only having issue with the Tesla K80.

Any help would be very appreciated.