Now I'm on Chieftec 850W. I know that it is too low power, but tomorrow will recieve Sirtec 1200W. I want to connect 4 gpu's just to know that port's are working and will start mining with all GPU's when I get Sirtec 1200W.
I want to run 5 x R9 380's on a EVGA 1300 but i don't know if that will be enought power for that amount of gpu's
Enough, but GPU undervolting is recommended. ETH algo is not power-hungry, but if you want someday to switch your rig to another mining algorithm , things may change.
I want to run 5 x R9 380's on a EVGA 1300 but i don't know if that will be enought power for that amount of gpu's
Enough, but GPU undervolting is recommended. ETH algo is not power-hungry, but if you want someday to switch your rig to another mining algorithm , things may change.
So, I reset the motherboard, I did again a clean install of the windows 10, installed the latest, latest 16.2 Crimson Driver. Connected via PCI-E x16 and x8 slots two cards, everything worked great. Then I connected the 3rd card in PCI-E x1 slot and I got blue screen with error: page_fault_in_nonpaged_area (atikmdag.sys). I found on the internet that this is AMD driver problem. Ok. The same card that I tried to connect via PCI-E x1 I connected with extender to x4 slot and windows recognized it, so I have three cards working. After that I turned off the computer and moved back card from x4 slot to x1 slot. And it worked! The computer recognized it! ( So I suppose that card when was connected via extender to x4 slot, installed all necessary updates and remained in computer memory. That's why now it was recognized ). But when I did the same thing with the fourth card, it did not work anymore and if I try to put back in slot PCI-E x1 the card that worked one time, it does not work anymore .
Now I mining with 4 GPU out of 5. One GPU is in my PC and another 3 in the rig I bought for mining ( the one I can't configure to work via PCI x1 ). As I started this Friday first mining in my life, I started to mine on Dwarfpool, because I can get reward faster, so I can test the system ( transfer from wallet to exchange, exchange service and withdrawals ). Today I moved to Ethpool, I want to compare them. My ethminer show me that ASUS R9 380 have 19,5 MH/s and 3 x MSI R9 380 have 60,5 MH/s, so total about 80 MH/s. But on Dwarfpool after 24 hours of mining, every hour Hash rate oscilated between 90-110 MH/s, Only once it dropped to 79 MH/s. I know that pool calculate MH/s by submited share in an amount of time. But in 24 Hours it was mostly always at 90-95 MH/s. What is the real number? My ethminer or average from pool? I will see after 12-24 hours, average on ethpool to compare ( and earnings ofcourse )
Today finally I received USB Risers and ...... it work just great. No more issue with PCI-E x1. It connects with no problem. So I suppose that flex risers I have are very bad quality! Because they worked once or twice and that's it For sure that flex risers have some problem connectors. If I would have an option to get USB risers quicker, I would resolve my problem very fast...but I live in Romania and here there are very few options to find USB Risers quick.
Some boards require what's called a PCI-E presence short in order to properly detect 1x risers in some slots, and the USB risers have the presence short built in.
And yes, ribbon risers break incredibly easily, I've broken one last 4 times I've worked on a rig using them, slowly replacing em all with the USB style that are far superior.
@dilux Yes, I have now MSI and Asus Strix R9 380 cards and all Work ok. My problem was only with risers and motherboard. Did you tried card one by one?
@dilux Now I remember that I had problem with same message for one of GPU's, but It was because of the PCI-E port. When I plugged It in another port It worked fine. So problem was in motherboard, not GPU.
@dilux Now I remember that I had problem with same message for one of GPU's, but It was because of the PCI-E port. When I plugged It in another port It worked fine. So problem was in motherboard, not GPU.
turns out the motherboard (gigabyte ga-z87x-oc) just didnt like these asus cards. i tried the same cards in an asrock h81 and it worked right away even without setting extra params
@verjic I have been having lots of issues with this GIGABYTE G1 Gaming GA-Z97X-Gaming 3 myself. I have finally gotten it to run all 6 cards, but after a while before now it kept freezing. Still testing, did you have any freezing issues? Did you edit BIOS settings for the PCIE settings (1x vs 4x), or turning off on mobo graphics? Trying to figure it out for 2+ weeks now.
Comments
Asus GPU Tweak recognize other cards, but it work best with ASUS cards.
@Masked_Immortal
Thank's. Will try today this plan.
I want to connect 4 gpu's just to know that port's are working and will start mining with all GPU's when I get Sirtec 1200W.
ETH algo is not power-hungry, but if you want someday to switch your rig to another mining algorithm , things may change.
So, I reset the motherboard, I did again a clean install of the windows 10, installed the latest, latest 16.2 Crimson Driver. Connected via PCI-E x16 and x8 slots two cards, everything worked great. Then I connected the 3rd card in PCI-E x1 slot and I got blue screen with error: page_fault_in_nonpaged_area (atikmdag.sys). I found on the internet that this is AMD driver problem. Ok. The same card that I tried to connect via PCI-E x1 I connected with extender to x4 slot and windows recognized it, so I have three cards working. After that I turned off the computer and moved back card from x4 slot to x1 slot. And it worked! The computer recognized it! ( So I suppose that card when was connected via extender to x4 slot, installed all necessary updates and remained in computer memory. That's why now it was recognized ). But when I did the same thing with the fourth card, it did not work anymore and if I try to put back in slot PCI-E x1 the card that worked one time, it does not work anymore
My ethminer show me that ASUS R9 380 have 19,5 MH/s and 3 x MSI R9 380 have 60,5 MH/s, so total about 80 MH/s. But on Dwarfpool after 24 hours of mining, every hour Hash rate oscilated between 90-110 MH/s, Only once it dropped to 79 MH/s. I know that pool calculate MH/s by submited share in an amount of time. But in 24 Hours it was mostly always at 90-95 MH/s. What is the real number? My ethminer or average from pool?
I will see after 12-24 hours, average on ethpool to compare ( and earnings ofcourse )
Today finally I received USB Risers and ...... it work just great. No more issue with PCI-E x1. It connects with no problem. So I suppose that flex risers I have are very bad quality! Because they worked once or twice and that's it
I want to thank all you guys @Jukebox @happytreefriends @Masked_Immortal @Logicaluser , for trying to help! I really appreciate your effort to resolve the "mystery" of PCI-E x1.
And as a conclusion better to wait and buy high quality USB Risers, than waste hours of your time and be angry all that time!!
Some boards require what's called a PCI-E presence short in order to properly detect 1x risers in some slots, and the USB risers have the presence short built in.
And yes, ribbon risers break incredibly easily, I've broken one last 4 times I've worked on a rig using them, slowly replacing em all with the USB style that are far superior.
i got 4 strix's but they wont mine because of can't allocate the dag in a single chunk. bailing.
did you try to enter this before mining : setx GPU_MAX_ALLOC_PERCENT 100
export GPU_MAX_ALLOC_PERCENT=100
export GPU_MAX_HEAP_SIZE=100
this seems to fix the problem for others that run into this problem with 2GB cards.. bot not for me
thx.
I realy need some advice; I'm trying to mine with 2 gigabyte r9 380 and my hashrate is too low! now is 32mh/s! any ideas??
thanks so much for the help!