Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux)

[telep]

Hi all, how are you?

I´m having a strange issue mining on claymore eth:

Since 2days ago, the miner program restarts every 5-10 minutes, with no gpu errors, speed dcrease on any gpu or something like that.
It only happens with claymore ethereum, 11.5,11.4 ...  Any problem with sgminer, claymore xmr or other miners.

This is a 10gpu rx470/570 rig, asrock h110 running SMOS under linux.


Any ideas?? Anyone had this issue before?

Virtual memory do you have 32-64GB of it?

Post your .bat batch file

Hi!
This is under linux, simplemining manages the virtualmemory but it always works without problem with this config.
Claymore´s command line :
-wd 1 -r 1 -epool stratum+tcp://daggerhashimoto.eu.nicehash.com:3353 -ewal "adress" -epsw x -esm 3 -allpools 1 -asm 1 -estale 0 -mport 127.0.0.1:3333

[1714027534]

1714027534

[1714027534]

1714027534

[1714027534]

1714027534

[1714027534]

1714027534

[polofitted007]

For the first 15 - 24 hrs, it has no problems, and then the miner hangs, hash zero until I hard shutdown the PC.

Setup...

GPUs: 6x 1060 6gbs Miner: Claymore 11.5 64bit OS: Windows 10 64 bit pro



Logs from right before the miner hangs...


02:02:45:424 1b04 ETH: Share accepted (141 ms)!

02:02:46:802 13c4 em hbt: 32, fm hbt: 125, 02:02:46:802 13c4 watchdog - thread 0 (gpu0), hb time 32 02:02:46:802 13c4 watchdog - thread 1 (gpu0), hb time 157 02:02:46:802 13c4 watchdog - thread 2 (gpu1), hb time 94 02:02:46:815 13c4 watchdog - thread 3 (gpu1), hb time 234 02:02:46:815 13c4 watchdog - thread 4 (gpu2), hb time 265 02:02:46:831 13c4 watchdog - thread 5 (gpu2), hb time 156 02:02:46:831 13c4 watchdog - thread 6 (gpu3), hb time 156 02:02:46:846 13c4 watchdog - thread 7 (gpu3), hb time 46 02:02:46:846 13c4 watchdog - thread 8 (gpu4), hb time 140 02:02:46:846 13c4 watchdog - thread 9 (gpu4), hb time 0 02:02:46:862 13c4 watchdog - thread 10 (gpu5), hb time 78 02:02:46:862 13c4 watchdog - thread 11 (gpu5), hb time 203 02:02:47:018 1b04 ETH: checking pool connection... 02:02:47:018 1b04 sent: {"worker": "", "jsonrpc": "2.0", "params": [], "id": 3, "method": "eth_getWork"}

02:02:47:190 1b04 buf: {"jsonrpc":"2.0","id":0,"result":["0x8af772e5207971a9c17739cb0bd444c77bb74ea0ea1515da9384a1f47483ff00","0x40910b6ac043b9a4c8a3a5100b00efcc1d90c39643cc754f24bf544c3e29c01c","0x000000006df37f675ef6eadf5ab9a2072d44268d97df837e6748956e5c6c2116"]}

02:02:53:456 1b04 sent: {"id":6,"jsonrpc":"2.0","method":"eth_submitHashrate","params":["0x763ca7b", "0x00000000000000000000000000000000000000000000000000000000029a3eb2"]}

02:02:54:145 16ac GPU0 t=69C fan=62%, GPU1 t=69C fan=34%, GPU2 t=68C fan=60%, GPU3 t=74C fan=42%, GPU4 t=83C fan=73%, GPU5 t=67C fan=55% 02:02:57:035 1b04 ETH: checking pool connection... 02:02:57:035 1b04 sent: {"worker": "", "jsonrpc": "2.0", "params": [], "id": 3, "method": "eth_getWork"}

02:02:57:191 1b04 buf: {"jsonrpc":"2.0","id":0,"result":["0x8af772e5207971a9c17739cb0bd444c77bb74ea0ea1515da9384a1f47483ff00","0x40910b6ac043b9a4c8a3a5100b00efcc1d90c39643cc754f24bf544c3e29c01c","0x000000006df37f675ef6eadf5ab9a2072d44268d97df837e6748956e5c6c2116"]}

02:03:01:457 1b04 buf: {"jsonrpc":"2.0","id":0,"result":["0xb768f375031755ac0898d84062207f4a494d761d850bf0bd2e5024cbd30fc4f7","0x40910b6ac043b9a4c8a3a5100b00efcc1d90c39643cc754f24bf544c3e29c01c","0x000000006df37f675ef6eadf5ab9a2072d44268d97df837e6748956e5c6c2116"]}

02:03:01:457 1b04 ETH: 03/29/18-02:03:01 - New job from eth-us-east1.nanopool.org:9999 02:03:01:473 1b04 target: 0x000000006df37f67 (diff: 10000MH), epoch 178(2.39GB) 02:03:01:473 1b04 ETH - Total Speed: 129.842 Mh/s, Total Shares: 219, Rejected: 0, Time: 04:31 02:03:01:492 1b04 ETH: GPU0 21.460 Mh/s, GPU1 21.693 Mh/s, GPU2 21.881 Mh/s, GPU3 21.703 Mh/s, GPU4 21.559 Mh/s, GPU5 21.546 Mh/s

Any help?

Here is my batch file...

setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100
EthDcrMiner64.exe -epool eth-us-east1.nanopool.org:9999 -ewal 0x6431245d999c72bd4cce8c1941e5410017/Sleepy/polofd007@yahoo.com -epsw x -ftime 10 -mode 1

[androstan1234]

For some reason, since the last windows update, i can no longer use the #0 card, it hangs out setting the dag.
Has anybody experienced this same problem?
I can't go back the windows version, so i'll move to linux (already have a lot of linux rigs so it's not a big problem) but i don't have the time right now, so i wanted to have it fully working for a few days more until the migration

Thanx guys!

Same thing happened to me.  Windows freezes when setting dag epoch on gpu0.  Problem is only on my TB85 rx580 rig with 4gb of system ram.  My TB250 rx570 rig with 8gb of system ram is fine.  Both are six-card rigs.

Work has been too busy so I haven't had time to troubleshoot.  Just used -di to turn off gpu0 for now.

If you figure something out plz let me know.

If anyone else has insight, please share.

To clarify, I did some very quick and rudimentary troubleshooting.  Ddu and clean install of aug23 blockchain driver.  Checked riser.  GPU0 appears normal in gpuz and msi AB.  No time atm to do much more.

I wonder if increasing virtual memory might help.  Or upgrading to 8GB system ram.

I don't want to just try something remotely bc if my rig freezes I won't be able to go to it in person to turn the PSU off/on for at least a few days.

[HELIUM_COIN]

Is it over for 3GB cards on ether?  

Getting CUDA error 11 - cannot write buffer for DAG.

Tried the -eres 0 and -lidag 0

Was mining fine for months until today.

Thanks!

[Claymore]

It seems 3GB cards won't be able to mine ETH/ETC in Windows10 soon because this OS reserves about 20% of GPU memory size and there is no way to fix it (at least easily).
In next update I will set zero devfee for these cards (same as I did for 2GB) so you can mine ETH forks.
If you still want to mine ETH/ETC on 3GB cards you should use Windows7 or Linux.

[Bigdrago]

For some reason, since the last windows update, i can no longer use the #0 card, it hangs out setting the dag.
Has anybody experienced this same problem?
I can't go back the windows version, so i'll move to linux (already have a lot of linux rigs so it's not a big problem) but i don't have the time right now, so i wanted to have it fully working for a few days more until the migration

Thanx guys!

Same thing happened to me.  Windows freezes when setting dag epoch on gpu0.  Problem is only on my TB85 rx580 rig with 4gb of system ram.  My TB250 rx570 rig with 8gb of system ram is fine.  Both are six-card rigs.

Work has been too busy so I haven't had time to troubleshoot.  Just used -di to turn off gpu0 for now.

If you figure something out plz let me know.

If anyone else has insight, please share.

To clarify, I did some very quick and rudimentary troubleshooting.  Ddu and clean install of aug23 blockchain driver.  Checked riser.  GPU0 appears normal in gpuz and msi AB.  No time atm to do much more.

I wonder if increasing virtual memory might help.  Or upgrading to 8GB system ram.

I don't want to just try something remotely bc if my rig freezes I won't be able to go to it in person to turn the PSU off/on for at least a few days.

Do u have a monitpr connected? Internal gpu enabled?

[freeapp]

ahh, finally, i can sleep. after long haul of testing.

been trying to get 11.5 stable, aiming 0 incorrect shares but i failed.

no matter how i downclock memory, add voltage, etc, still no luck. hwinfo shows 0 memory error for 24hrs, still, 11.5 keeps sprouting incorrect shares at random gpus. it leads me to the point that im almost losing 3-4mh/s per card. meh, more like stock settings but with incorrect shares.

then i finally downgraded to v11.0 (using -asm 2 btw as it yields me better hashrate)

reverted my old stable OC

now im enjoying 0 incorrect shares for 24hrs+, more hashrate, and improved effective pool hashrate.

here are my specs btw:
rig1: 13pcs rx570 red devil (hynix;elpida) 386mh/s(with 11.0)
rig2: 13pcs rx570 nitro+ 4gb (samsung) 395mh/s(with 11.0)
os: win10 1709
driver: adrenalin 18.3.4
(with 11.5, i almost downgraded to 360mh/s per rig, still gives me incorrect shares at random gpus)

with 11.0, i only get ~1mh/s total rig fluctuation

with 11.5, even though using -oldkernels, i get big fluctuations, around 20mh/s total fluctuation and it takes too long to revert to original hashrate.

11.5 gives flat reported hashrate with new kernels, but yeah, cant make it stable without sprouting incorrect shares.

so thats all about it. i dont blame no miners, but its my personal experience.

some doesnt care much about incorrect shares, but i do, it translates to wasted hashrate.

i believe, problems to me started at v11.1 and beyond.

It could be your Bios mod. I'm running v11.5 with the Adrenaline v18.3.4 drivers on two rigs and it's very stable with zero incorrect or invalid shares. I even increased my overclock on my ASRock H110 Pro BTC+ RX 480/580 13 card rig and got an extra 8 MH/s. On the 6 card Biostar TB85 rig I got an extra 4 MH/s.

I'm using the Polaris Bios Editor 'One Click Timing Patch' for all my cards. The only change to the PBE bundled performance timings I made was replace the Samsung UberMix v3.1 timing, which is too aggresive with this one that is much more stable.

Code:

777000000000000022CC1C00106A5D4DD0571016B90D060C0060070014051420FA8900A0030000001011333DC0303A17

https://github.com/jaschaknack/PolarisBiosEditor


-ethi 8
-cclock 1200,1150,1200,1200,1200,1200,1200,1200,1200,1200,1200,1200,1200
-mclock 2050,2100,2100,2250,2150,2150,2150,2250,2225,2250,2150,2250,2100
-powlim -10,-5,-10,-15,-10,-15,-10,-10,-10,-10,-10,-10,-10
-cvddc 950,950,950,850,900,850,850,900,900,900,850,900,900
-mvddc 950,950,950,850,900,850,850,900,900,900,850,900,900


-ethi 8
-cclock 1150,1150,1200,1150,1200,1200
-mclock 2075,2000,2075,2075,2075,1950
-powlim -17,-10,-10,-15,-15,-15
-cvddc 925,925,925,925,900,925
-mvddc 925,925,925,925,900,925

Hi,
 i have just modded my bios and uploaded it again.
But Windows told me that there is a problem with this card so it can't be activated.
Just googled a bit, does i have to install  *atikmdag-patcher* after driver installation? If yes is there a newer version instead of 1.4.6 available? I have also installed 18.3.4...

thx in advance.

best regards,
freeapp


EDIT:
I have found the answer by myself...
Yes i have to install the *atikmdag-patcher* ...

Thx for everyone who has read this and wanted to help...

[HELIUM_COIN]

Thanks Claymore!

It seems 3GB cards won't be able to mine ETH/ETC in Windows10 soon because this OS reserves about 20% of GPU memory size and there is no way to fix it (at least easily).
In next update I will set zero devfee for these cards (same as I did for 2GB) so you can mine ETH forks.
If you still want to mine ETH/ETC on 3GB cards you should use Windows7 or Linux.

[deadsix]

It seems 3GB cards won't be able to mine ETH/ETC in Windows10 soon because this OS reserves about 20% of GPU memory size and there is no way to fix it (at least easily).
In next update I will set zero devfee for these cards (same as I did for 2GB) so you can mine ETH forks.
If you still want to mine ETH/ETC on 3GB cards you should use Windows7 or Linux.

Great. So No devfee on 3GB cards for both windows and linux versions?

[Vann]

It seems 3GB cards won't be able to mine ETH/ETC in Windows10 soon because this OS reserves about 20% of GPU memory size and there is no way to fix it (at least easily).
In next update I will set zero devfee for these cards (same as I did for 2GB) so you can mine ETH forks.
If you still want to mine ETH/ETC on 3GB cards you should use Windows7 or Linux. Windows

Since Windows 7 is limited to 4 cards, I think Windows Server 2016 is another option for using Windows if you have more than 4 cards. There is also a 180-day evaluation of Windows Server 2016 for free.

https://www.microsoft.com/en-us/evalcenter/evaluate-windows-server-2016

[Viaggiatore]

Hello, not properly related to Claymore I think but... second time it happens to me, my 8 GPU rig seems "freezing", it's not answering to ping nor to TeamViewer. This for say 5 minutes then it seems it wake up by itself. Of course I have sleep disabled! Also updates are disabled! I had a look at logs but nothing special is there??? Any hint on how to find out what could possibly be? Where to look at in logs?

Thanks!

[Vann]

Try setting the Windows power plan to 'high performance'.

[Viaggiatore]

Try setting the Windows power plan to 'high performance'.

Thanks, but my rig is definetely not going to sleep. It seems to me that just internet connections is dropped (I have both ethernet and wifi for redundancy). Seems both of them are killed for apparently no reason...

[columbiamining]

@Viaggiatore Your rig is probably blue screening and restarting. Check Windows Event Viewer for logs. I'd suggest DDU and reinstall latest 18.3.4 drivers.

[Vann]

Try setting the Windows power plan to 'high performance'.

Thanks, but my rig is definetely not going to sleep. It seems to me that just internet connections is dropped (I have both ethernet and wifi for redundancy). Seems both of them are killed for apparently no reason...

It doesn't have to go to sleep. Your power plan power management mode could be setting the network adapters to power savings mode, which is why you should set the power plan to high performance. If you want to do it manually, under power options select 'change advanced power settings' and verify the network adapters are set to 'maximum performance'.

[Viaggiatore]

@Viaggiatore Your rig is probably blue screening and restarting. Check Windows Event Viewer for logs. I'd suggest DDU and reinstall latest 18.3.4 drivers.

Thanks, but definitely it's not the case, in Claymore window I can see it is just the internet connection which went bananas ;-)

I tried with unchecking "Allow the computer to turn off this device to save power" on both ethernet and wifi device, I read somewhere this could be the issue.
I hope it will not break WOL btw

[Viaggiatore]

Try setting the Windows power plan to 'high performance'.

Thanks, but my rig is definetely not going to sleep. It seems to me that just internet connections is dropped (I have both ethernet and wifi for redundancy). Seems both of them are killed for apparently no reason...

It doesn't have to go to sleep. Your power plan power management mode could be setting the network adapters to power savings mode, which is why you should set the power plan to high performance. If you want to do it manually, under power options select 'change advanced power settings' and verify the network adapters are set to 'maximum performance'.

It was already in "maximum performance". BTW listened to you and now set to high performance, also unchecked "Allow the computer to turn off this device to save power" on both the network devices, hoping it won't break WOL/WOW

[androstan1234]

For some reason, since the last windows update, i can no longer use the #0 card, it hangs out setting the dag.
Has anybody experienced this same problem?
I can't go back the windows version, so i'll move to linux (already have a lot of linux rigs so it's not a big problem) but i don't have the time right now, so i wanted to have it fully working for a few days more until the migration

Thanx guys!

Same thing happened to me.  Windows freezes when setting dag epoch on gpu0.  Problem is only on my TB85 rx580 rig with 4gb of system ram.  My TB250 rx570 rig with 8gb of system ram is fine.  Both are six-card rigs.

Work has been too busy so I haven't had time to troubleshoot.  Just used -di to turn off gpu0 for now.

If you figure something out plz let me know.

If anyone else has insight, please share.

To clarify, I did some very quick and rudimentary troubleshooting.  Ddu and clean install of aug23 blockchain driver.  Checked riser.  GPU0 appears normal in gpuz and msi AB.  No time atm to do much more.

I wonder if increasing virtual memory might help.  Or upgrading to 8GB system ram.

I don't want to just try something remotely bc if my rig freezes I won't be able to go to it in person to turn the PSU off/on for at least a few days.

Do u have a monitpr connected? Internal gpu enabled?

Yes igpu enabled and dummy plug connected to mobo dvid.

[awdmesh]

Never did fix my instability issue with 11.5 so I rolled back to 10.6 and it just churns away, except for the other night when I quickly realized the one 3GB card I have amongst the other 6 AMD cards wasn't creating the DAG. Some quick research and it of course led me to what Claymore mentioned.

I will say, setting -eres 0 got it back up and mining but correct me if i'm wrong, each time the DAG changes/increases I'll get a miner/machine reboot? At least that's what's happened so far when it jumped from 177 to 178 or whatever it was. I'm assuming that's because the cards are not reserving the space. Looks like I'm going to have to swap out a 470 for the 1060.

[tg88]

Never did fix my instability issue with 11.5 so I rolled back to 10.6 and it just churns away, except for the other night when I quickly realized the one 3GB card I have amongst the other 6 AMD cards wasn't creating the DAG. Some quick research and it of course led me to what Claymore mentioned.

What amd driver version are you using?

for me latest 18.3.4 drivers solved the instability..