fullzero (OP)
Legendary
Offline
Activity: 1260
Merit: 1009
|
|
August 13, 2017, 04:39:27 PM |
|
Hi guys,
how can we add this command on bash file?
GPU_FORCE_64BIT_PTR 0 GPU_MAX_HEAP_SIZE 100 GPU_USE_SYNC_OBJECTS 1 GPU_MAX_ALLOC_PERCENT 100 GPU_SINGLE_ALLOC_PERCENT 100
thanks
God bless
Hi, Like this export GPU_FORCE_64BIT_PTR="0" export GPU_MAX_HEAP_SIZE="100" export GPU_USE_SYNC_OBJECTS="1" export GPU_MAX_ALLOC_PERCENT="100" export GPU_SINGLE_ALLOC_PERCENT="100" set: USE_ENVIRONMENTAL_VARIBLES="YES"
|
|
|
|
JayneL
Member
Offline
Activity: 104
Merit: 10
|
|
August 13, 2017, 05:35:42 PM |
|
Opppss thanks got it.. Guys can you share ur 1050ti OC as I still tweaking to get the best stable OC for my 12 1050ti palit/inno3d in 1bash change USE_ENVIRONMENTAL_VARIBLES from "NO" to "YES" USE_ENVIRONMENTAL_VARIBLES="YES" #YES NO they have some code below aready in 1bash if [ $USE_ENVIRONMENTAL_VARIBLES == "YES" ] then export GPU_FORCE_64BIT_PTR="0" export GPU_MAX_HEAP_SIZE="100" export GPU_USE_SYNC_OBJECTS="1" export GPU_MAX_ALLOC_PERCENT="100" echo "" echo ENVIRONMENTAL VARIBLES SET fi Hi guys,
how can we add this command on bash file?
GPU_FORCE_64BIT_PTR 0 GPU_MAX_HEAP_SIZE 100 GPU_USE_SYNC_OBJECTS 1 GPU_MAX_ALLOC_PERCENT 100 GPU_SINGLE_ALLOC_PERCENT 100
thanks
God bless
|
|
|
|
Bibi187
Full Member
Offline
Activity: 420
Merit: 106
https://steemit.com/@bibi187
|
|
August 13, 2017, 05:49:19 PM |
|
This setting is not only for AMD ?
GPU_FORCE_64BIT_PTR 0 GPU_MAX_HEAP_SIZE 100 GPU_USE_SYNC_OBJECTS 1 GPU_MAX_ALLOC_PERCENT 100 GPU_SINGLE_ALLOC_PERCENT 100
Wath the point to put on YES on env setting when u use nvidia ? I prolly miss something my RIG run fine with this option disable
|
|
|
|
lards
Newbie
Offline
Activity: 50
Merit: 0
|
|
August 13, 2017, 06:19:56 PM |
|
IAmNotAJeep Thought that quoting everything gonna create a mess
Have you been using 2 ATX PSUs or one was server PSU like in my set up? I don't have quick access to new PSUs so swapping them gonna take about 5 days to one week , it's very strange how this error seems to appear in completely unpredictable timings. I also worked out strange thing that when I disconnect the PSUs from each other it seems more stable than with the rele connector synchronizing them , when they are connected it takes a couple of minutes to freeze and otherwise it can work for a couple of hours even with more then 8 cards( using 11 atm seems to be impossible for some reason to plug in all 12 without something crashing).
I have been building new rigs and troubleshooting existing ones in a mine. I would try IAmNotAJeep's suggestion; I have also had dual joined PSUs cause problems before. I have seen many similar problems (even with 6 gpu rigs) and then also had them disappear when I remove all the trouble GPUs and make another rig out of them. This has been the most time effective way of resolving all GPU problems; however, it does have a high cost in regard to using additional components. Thanks fullzero for your reply. I am going to find parts and cables to add 2 molexes to my server PSU and try to use single PSU and see how that works out as now it seems okayish with 11 cards , just the 12th one causing lots of trouble with drivers and the system overall. Also as I have 11 P106-100 cards available and you can probably if you need connect to my rig to try and see what the problem in OC is , teamviewer or what do you need to use? Would highly appreciate you having a solution for this since these cards can run up to 25 MH/s and now all I can get from them is 21 which is 15% difference and since I was thinking of making over 5 of those rigs this makes huge impact on profitability. If you can somehow deal with OC issue on P106-100 I think me as well as other users using those cards wouldn't hesitate on making good donation for your work since this system is probably one of the best if not the best for Nvidia Cards atm.
|
|
|
|
mraug
Newbie
Offline
Activity: 25
Merit: 0
|
|
August 13, 2017, 10:23:13 PM |
|
Got a rig with 6x1080ti, mining ZEC Was doing fine for about 2 weeks but now GPU4 stops: https://i.imgur.com/iXJoYLH.jpgWatchdog restarts the mining process but GPU4 doesn't recover. https://i.imgur.com/oorGsXW.jpgTeam viewer and miner freezes when I try to access the Guake terminal, but the rig is still accessible via SSH and seems like the RSS process also still works because the rig is not rebooted. I already replaced raiser, didn't help. I dropped OC down to ZERO both on GPU and Memory, didn't help. I raised POWER from 180 to 230, no effect on stability. Any ideas how to troubleshoot? try lower power say 160 or pull the bad card run it on a tester mobo that runs troubled cards. Your tester mobo needs to be able to run 2 or 3 cards not six and if you don't have a tester mobo buy a cheap one like this https://www.newegg.com/Product/Product.aspx?Item=N82E16813128980Found the problem, it was related to the PSU connector board: http://www.ebay.com/itm/Desktop-ATX-24Pin-Dual-PSU-Power-Supply-Sync-Starter-Extender-Cable-Connector-/302285871050?epid=22006157212&hash=item4661a457ca:g:VxgAAOSwj25Y8cz6All my rigs are dual PSU and all of them worked fine. Just this one started to act up after 2 weeks. I removed all cards from one of the PSUs and the rig crashed anyway, at that point it was clear to me it must have something to do with the PSU connector. I replaced it and it's been running now for 2-3 days. Seems to me it's a common problem, looking at the recent posts of IAmNotAJeep.
|
|
|
|
car1999
|
|
August 14, 2017, 12:25:01 AM |
|
I see these codes in 1bash: if [ $P106_100 == "YES" ] then XORG="OK" fi
Does that mean I can setup a rig with all of p106-100 cards and connect monitor to the motherboard directly? In theory yes; although I haven't tested this myself. Although the P106_100's will also likely not OC. I really need 2x P106_100's to test. I tried to connect to ivoldemar's rig with them yesterday; but my connection was refused. I'll try again today. my p106*8 rig will be ready in 2 days, you can ssh to it to do some tests.
|
|
|
|
JayneL
Member
Offline
Activity: 104
Merit: 10
|
|
August 14, 2017, 07:58:06 AM |
|
can you share your stable OC for 1050ti in ETH / DUAL_ETH_SC i cant make it to run for a straight 10hrs it alwahs have a random hang
my oc Cclock : 150 Mclock : 1400 Powlimit : 58
please anyone got same issue?
thanks
God Bless
|
|
|
|
lards
Newbie
Offline
Activity: 50
Merit: 0
|
|
August 14, 2017, 06:35:55 PM Last edit: August 14, 2017, 09:55:51 PM by lards |
|
Hi fullzero.
I've come across another problem related to p106-100 cards which is the following :
Been running 11 cards on h110 asrock and then tried to connect 12th one and everytime I do it Nvidia drivers seem to crash no matter what I do. It's not a hardware problem as I have tested this card in 11 cards set up and everything boots fine, just the 12th one seem to crash everything and I have to reflash the OS again to make it work. When I try with fresh OS and 12 cards its the same problem again, the nvidia-smi file isn't found and the claymore can't initialize CUDA drivers .
Has anyone seen this before and maybe knows the way to deal with this? Maybe I need to correct something in 1bash to make it work but so far all my tries were a failure and the miner couldn't find drivers.
Thanks for your help.
UPD. Somehow managed to get all 12 cards working but in genoil instead of claymore , still have no idea what the problem was since I've done everything like swapping risers and cards before and this hasn't got me anywhere, but genoil gives lower hashrate then claymore, so does anyone know or can advice on how to make claymore work aat this set up?
|
|
|
|
gs777
Member
Offline
Activity: 118
Merit: 10
|
|
August 14, 2017, 10:00:35 PM |
|
Hi fullzero.
I've come across another problem related to p106-100 cards which is the following :
Been running 11 cards on h110 asrock and then tried to connect 12th one and everytime I do it Nvidia drivers seem to crash no matter what I do. It's not a hardware problem as I have tested this card in 11 cards set up and everything boots fine, just the 12th one seem to crash everything and I have to reflash the OS again to make it work. When I try with fresh OS and 12 cards its the same problem again, the nvidia-smi file isn't found and the claymore can't initialize CUDA drivers .
Has anyone seen this before and maybe knows the way to deal with this? Maybe I need to correct something in 1bash to make it work but so far all my tries were a failure and the miner couldn't find drivers.
Thanks for your help.
Don't know is that's the problem or not but I hope it'l help H110 Pro BTC+ has 5 power connectors ATX power 1 (4) ATX 12V (1) MOLEX 1 (21) MOLEX 2 (19) SATA power 1 (18) ASRock in their manual (page 20) suggest to connect them to power supply http://asrock.pc.cdn.bitgravity.com/Manual/H110%20Pro%20BTC+.pdf
|
|
|
|
scryptr
Legendary
Offline
Activity: 1797
Merit: 1028
|
|
August 14, 2017, 10:04:23 PM |
|
This setting is not only for AMD ?
GPU_FORCE_64BIT_PTR 0 GPU_MAX_HEAP_SIZE 100 GPU_USE_SYNC_OBJECTS 1 GPU_MAX_ALLOC_PERCENT 100 GPU_SINGLE_ALLOC_PERCENT 100
Wath the point to put on YES on env setting when u use nvidia ? I prolly miss something my RIG run fine with this option disable
THE SET OF ENVIRONMENTAL VARIABLE VALUES IS FOR AMD GRAPHICS CARDS-- This block of environmental variable commands is AMD Radeon specific. The Green Team (nVidia) does not need them. --scryptra
|
|
|
|
lards
Newbie
Offline
Activity: 50
Merit: 0
|
|
August 14, 2017, 10:15:00 PM |
|
Hi fullzero.
I've come across another problem related to p106-100 cards which is the following :
Been running 11 cards on h110 asrock and then tried to connect 12th one and everytime I do it Nvidia drivers seem to crash no matter what I do. It's not a hardware problem as I have tested this card in 11 cards set up and everything boots fine, just the 12th one seem to crash everything and I have to reflash the OS again to make it work. When I try with fresh OS and 12 cards its the same problem again, the nvidia-smi file isn't found and the claymore can't initialize CUDA drivers .
Has anyone seen this before and maybe knows the way to deal with this? Maybe I need to correct something in 1bash to make it work but so far all my tries were a failure and the miner couldn't find drivers.
Thanks for your help.
Don't know is that's the problem or not but I hope it'l help H110 Pro BTC+ has 5 power connectors ATX power 1 (4) ATX 12V (1) MOLEX 1 (21) MOLEX 2 (19) SATA power 1 (18) ASRock in their manual (page 20) suggest to connect them to power supply http://asrock.pc.cdn.bitgravity.com/Manual/H110%20Pro%20BTC+.pdfYeah I know, all 5 are connected to the ATX PSU and risers+cards are powered from another PSU.
|
|
|
|
TenaciousJ
|
|
August 15, 2017, 08:26:50 AM |
|
Hey FZ,
I ran into an issue that I can't figure out how to fix. I've got 12 gpus on a TB250-BTC Pro, and all of them work fine, except the last one GPU12 (PCI:12:0:0) which doesn't seem to start an xserver instance on boot. Because it doesn't load xserver, the fan and power settings are unavailable for editing, so the card runs too hot and causes power throttling, and occasionally a reboot.
Any idea how to fix that? I checked the /etc/X11/xorg.conf file and it looks fine from what I can tell (all display device settings are there and look the same).
Thanks for your help.
|
|
|
|
ivoldemar
Newbie
Offline
Activity: 23
Merit: 0
|
|
August 15, 2017, 09:54:12 AM |
|
Hey, fullzero!
Tell me, when do you fix for P106 cards? I can not collect 40 farms already for 13 cards, because Does not work razgon on your assembly .... please give it 1 hour. I already threw you access to the rig on which there are 13 P106 cards. Access to you in private messages.
|
|
|
|
philipma1957
Legendary
Offline
Activity: 4242
Merit: 8497
'The right to privacy matters'
|
|
August 15, 2017, 03:21:13 PM |
|
okay nvOC 0018 has so many changes and alterations to it I can't follow them all.
I can get nvOC 0017 to work and have been stick with it.
hopefully nvOC 0019 comes out soon and it is easier to get it to work then nvOC 0018
at the moment I am running windows simply to get skunk to run on nicehash
I would prefer to run linux nvOC 0019 and have it as easy to set up as 0017 was.
I will await nvOC 0019 as I have done nvOC 0018 four times four fails.
|
|
|
|
Aravind_NAV
Newbie
Offline
Activity: 3
Merit: 0
|
|
August 15, 2017, 04:16:04 PM |
|
okay nvOC 0018 has so many changes and alterations to it I can't follow them all.
at the moment I am running windows simply to get skunk to run on nicehash
I would prefer to run linux nvOC 0019 and have it as easy to set up as 0017 was.
I (barely) managed to run the skunk algo on nvOC (on suprnova pool)... Here's basically what I did 1) Install openssl 1.0.1 (Sorry I'm new to this forum so I don't know how to link to other answers, but it's post #2515 on this thread by darklow 2) I used a different ccminer (ccminer 2.2) downloaded from github but after that, all commands worked like ./autogen.sh ./configure.sh ./build.sh (or make / make install) 3) Change 1bash file (Username, miner location, pool address etc. etc.) (Oh, and also I think it is best if you do all the steps in the same computer which runs the GPUs, not sure why though ) Hope I helped ! (This is my first 'helpful' post in bitcointalk.... Cheers !!!)
|
|
|
|
fullzero (OP)
Legendary
Offline
Activity: 1260
Merit: 1009
|
|
August 15, 2017, 05:01:53 PM |
|
This setting is not only for AMD ?
GPU_FORCE_64BIT_PTR 0 GPU_MAX_HEAP_SIZE 100 GPU_USE_SYNC_OBJECTS 1 GPU_MAX_ALLOC_PERCENT 100 GPU_SINGLE_ALLOC_PERCENT 100
Wath the point to put on YES on env setting when u use nvidia ? I prolly miss something my RIG run fine with this option disable
A member requested these as an option so I added them. I don't use them; although I tend to stick to the main coins. It is possible they make a difference when mining some of the more obscure coins.
|
|
|
|
fullzero (OP)
Legendary
Offline
Activity: 1260
Merit: 1009
|
|
August 15, 2017, 05:03:27 PM |
|
IAmNotAJeep Thought that quoting everything gonna create a mess
Have you been using 2 ATX PSUs or one was server PSU like in my set up? I don't have quick access to new PSUs so swapping them gonna take about 5 days to one week , it's very strange how this error seems to appear in completely unpredictable timings. I also worked out strange thing that when I disconnect the PSUs from each other it seems more stable than with the rele connector synchronizing them , when they are connected it takes a couple of minutes to freeze and otherwise it can work for a couple of hours even with more then 8 cards( using 11 atm seems to be impossible for some reason to plug in all 12 without something crashing).
I have been building new rigs and troubleshooting existing ones in a mine. I would try IAmNotAJeep's suggestion; I have also had dual joined PSUs cause problems before. I have seen many similar problems (even with 6 gpu rigs) and then also had them disappear when I remove all the trouble GPUs and make another rig out of them. This has been the most time effective way of resolving all GPU problems; however, it does have a high cost in regard to using additional components. Thanks fullzero for your reply. I am going to find parts and cables to add 2 molexes to my server PSU and try to use single PSU and see how that works out as now it seems okayish with 11 cards , just the 12th one causing lots of trouble with drivers and the system overall. Also as I have 11 P106-100 cards available and you can probably if you need connect to my rig to try and see what the problem in OC is , teamviewer or what do you need to use? Would highly appreciate you having a solution for this since these cards can run up to 25 MH/s and now all I can get from them is 21 which is 15% difference and since I was thinking of making over 5 of those rigs this makes huge impact on profitability. If you can somehow deal with OC issue on P106-100 I think me as well as other users using those cards wouldn't hesitate on making good donation for your work since this system is probably one of the best if not the best for Nvidia Cards atm. pm me teamviewer info for a P106-100 rig and I will try to get it working.
|
|
|
|
fullzero (OP)
Legendary
Offline
Activity: 1260
Merit: 1009
|
|
August 15, 2017, 05:04:37 PM |
|
I see these codes in 1bash: if [ $P106_100 == "YES" ] then XORG="OK" fi
Does that mean I can setup a rig with all of p106-100 cards and connect monitor to the motherboard directly? In theory yes; although I haven't tested this myself. Although the P106_100's will also likely not OC. I really need 2x P106_100's to test. I tried to connect to ivoldemar's rig with them yesterday; but my connection was refused. I'll try again today. my p106*8 rig will be ready in 2 days, you can ssh to it to do some tests. pm me teamviewer or ssh info for a P106-100 rig and I will try to get it working.
|
|
|
|
fullzero (OP)
Legendary
Offline
Activity: 1260
Merit: 1009
|
|
August 15, 2017, 05:06:03 PM |
|
can you share your stable OC for 1050ti in ETH / DUAL_ETH_SC i cant make it to run for a straight 10hrs it alwahs have a random hang
my oc Cclock : 150 Mclock : 1400 Powlimit : 58
please anyone got same issue?
thanks
God Bless
My guess is you need more power for stable dual mining. I recommend trying a powerlimit of 68 when dual mining with a 1050ti.
|
|
|
|
fullzero (OP)
Legendary
Offline
Activity: 1260
Merit: 1009
|
|
August 15, 2017, 05:08:41 PM |
|
Hi fullzero.
I've come across another problem related to p106-100 cards which is the following :
Been running 11 cards on h110 asrock and then tried to connect 12th one and everytime I do it Nvidia drivers seem to crash no matter what I do. It's not a hardware problem as I have tested this card in 11 cards set up and everything boots fine, just the 12th one seem to crash everything and I have to reflash the OS again to make it work. When I try with fresh OS and 12 cards its the same problem again, the nvidia-smi file isn't found and the claymore can't initialize CUDA drivers .
Has anyone seen this before and maybe knows the way to deal with this? Maybe I need to correct something in 1bash to make it work but so far all my tries were a failure and the miner couldn't find drivers.
Thanks for your help.
UPD. Somehow managed to get all 12 cards working but in genoil instead of claymore , still have no idea what the problem was since I've done everything like swapping risers and cards before and this hasn't got me anywhere, but genoil gives lower hashrate then claymore, so does anyone know or can advice on how to make claymore work aat this set up?
I'm guessing the 12th is assigned an odd index variable by the system that the xorg.conf or other system file doesn't properly deal with.
|
|
|
|
|