Topic: Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux) - page 310. (Read 6590718 times)
March 18, 2018, 07:39:25 AM
Excellent work! Stale share reduced! I'm dual mining, with -dcri 50, hashrate is increased for both. Thanx again Claymore!!!!
March 18, 2018, 07:12:13 AM
Truly odd one:
As of about two hours ago, 100% of my shares are being reported as stale..I am using 11.5 and had at most 40% ever, with fewer since 11.4/11.5 updates until then.
Niw I am actually seeing more reported stale than submitted, like the 9 or 10 sales with 8 total for a single card? Weird, but consistent for nearly 2 1/2 hours now.
I am single mining on ethermine.org
PS: (Post Mortem)
As always, I should have done a bit more research on things. I just got my first cup of coffee and didn't look at basic things like the reported share acceptance times. Watching more closely for a few minutes showed numbers like 5000ms or more than 5 full seconds.
I still have not pinned down exactly why it was, but since the rig had been running Windoze 10 for 18 days, I used the reboot method of fixing the problem and the times went back to being consistently under a second.
For me out in the middle of the national forest 20 miles from the closest true high speed internet, I'm on a satellite link, so the fact that I don't get under 50ms is normal. In fact 500ms is low for me and 750ms not unusual due to the propagation delay inherent to satellite up links and downlinks.
As of about two hours ago, 100% of my shares are being reported as stale..I am using 11.5 and had at most 40% ever, with fewer since 11.4/11.5 updates until then.
Niw I am actually seeing more reported stale than submitted, like the 9 or 10 sales with 8 total for a single card? Weird, but consistent for nearly 2 1/2 hours now.
I am single mining on ethermine.org
PS: (Post Mortem)
As always, I should have done a bit more research on things. I just got my first cup of coffee and didn't look at basic things like the reported share acceptance times. Watching more closely for a few minutes showed numbers like 5000ms or more than 5 full seconds.
I still have not pinned down exactly why it was, but since the rig had been running Windoze 10 for 18 days, I used the reboot method of fixing the problem and the times went back to being consistently under a second.
For me out in the middle of the national forest 20 miles from the closest true high speed internet, I'm on a satellite link, so the fact that I don't get under 50ms is normal. In fact 500ms is low for me and 750ms not unusual due to the propagation delay inherent to satellite up links and downlinks.
March 18, 2018, 06:39:41 AM
Claymore - Would it be possible to have the logs go into a subfolder in the next version?
+Or you could take a look at the readme and use "-logfile logs/logfile.txt"
Yes, also in next update I will improve "-logfile" option so you can specify folder for logs if last character is slash, for example "-logfile logs\".
March 18, 2018, 06:27:07 AM
Claymore - Would it be possible to have the logs go into a subfolder in the next version?
+Or you could take a look at the readme and use "-logfile logs/logfile.txt"
March 18, 2018, 06:04:22 AM
Claymore - Would it be possible to have the logs go into a subfolder in the next version?
+ March 18, 2018, 05:44:39 AM
Hi all
after upgrade to 11.5 from 11.2 I have lots of Incorrect ETH shares .. should I use oldkernel =1
after upgrade to 11.5 from 11.2 I have lots of Incorrect ETH shares .. should I use oldkernel =1
March 18, 2018, 05:07:27 AM
Claymore - Would it be possible to have the logs go into a subfolder in the next version?
+1. I love logs but they just create a mess in miners folder.
March 18, 2018, 04:36:28 AM
If anyone has any insight on this, been reading/searching for some time and cannot figure it out:
Have a single rig with 4x1050 TI and 1xRX 560 (yes, I'm small time - but sorry, already taken, ladies). Running v11.4 as 11.5 keeps locking the PC up completely. AMD drivers are 17.12.1. Mining in ETH only mode. Zero hardware errors on HWInfo and no BIOS/clock/voltage/power modifications on the AMD.
The AMD card is showing a hashrate of ~9.18 vs ~13.5 on all the Nvidia cards. Miner showing 0 rejected/incorrect shares after 24 hours of mining. AMD card has right at 50% of the average shares of the 4 Nvidias - on nanopool if that matters. Looking at the submissions from the logs and all the shares from the Nvidia cards are accepted within ~170ms, but the AMD are ~2200ms?
Trying to understand the variance on the AMD - if I go by the hash rate vs the Nvidias, it should be closer to 68% not 50%, and what is the deal with the shares taking so long to be accepted from this one card? Thanks
I have 560 and it gives me 14 m/hs with 1975 memory clock. Have a single rig with 4x1050 TI and 1xRX 560 (yes, I'm small time - but sorry, already taken, ladies). Running v11.4 as 11.5 keeps locking the PC up completely. AMD drivers are 17.12.1. Mining in ETH only mode. Zero hardware errors on HWInfo and no BIOS/clock/voltage/power modifications on the AMD.
The AMD card is showing a hashrate of ~9.18 vs ~13.5 on all the Nvidia cards. Miner showing 0 rejected/incorrect shares after 24 hours of mining. AMD card has right at 50% of the average shares of the 4 Nvidias - on nanopool if that matters. Looking at the submissions from the logs and all the shares from the Nvidia cards are accepted within ~170ms, but the AMD are ~2200ms?
Trying to understand the variance on the AMD - if I go by the hash rate vs the Nvidias, it should be closer to 68% not 50%, and what is the deal with the shares taking so long to be accepted from this one card? Thanks
March 18, 2018, 04:17:47 AM
Can someone please help or advise
Claymore doesn't show does not GPU temperature and fan status so I can't use fan and temrature related controls. Commands like -tt, -ttdcr, -ttli, -tstop, -tstart, -fanmax, -fanmin
RX 580 pulse 8 GB GPUs
Driver version 23.20.782.0
Driver date 20/10/2017
Radeon-Software-Crimson-Relive-Edition-17.10.2
MotherBoard B250 MINING EXPERT
Windows 10 ent edition x64
I am able to control fan and temprature via MSI afterburner
Claymore doesn't show does not GPU temperature and fan status so I can't use fan and temrature related controls. Commands like -tt, -ttdcr, -ttli, -tstop, -tstart, -fanmax, -fanmin
RX 580 pulse 8 GB GPUs
Driver version 23.20.782.0
Driver date 20/10/2017
Radeon-Software-Crimson-Relive-Edition-17.10.2
MotherBoard B250 MINING EXPERT
Windows 10 ent edition x64
I am able to control fan and temprature via MSI afterburner
UPDATE
Updated to radeon version 18.2.1, using Compute mode now. Good thing that hashrate gone up by 2 Mh/s each card.
temperature and fan stats started showing fine until rig crashed.
and now after the reboot same issue - can't see fan and temperature stats. seems like some sort of bug in claymore.
Can someone advise if they have seen the problem?
Has someone seen the issue? Is there some setting in driver?
Update 2
Issue: temperatures and fan settings break with latest amd drivers once rig crashes for some reason (example power outage)
Finally found the issue listed here https://bitcointalk.org/index.php?topic=2574589.40
it's AMD drivers causing the issue, claymore needs to update the code to get over the issue. It's pain in nack that claymore can't control fan, temp or OC settings
March 18, 2018, 03:59:09 AM
how to set in claymore to start up with windows on restart ?
Press windows button+R,type shell:startup and put there label of your bat file
March 18, 2018, 03:57:33 AM
how to set in claymore to start up with windows on restart ?
add your .bat file shortcut to your startup folder or directory
if your using win10
see this
https://www.howtogeek.com/208224/how-to-add-programs-files-and-folders-to-system-startup-in-windows-8.1/
March 18, 2018, 03:54:45 AM
how to set in claymore to start up with windows on restart ?
March 18, 2018, 03:44:23 AM
What does the -ethi parameter do? How does he work?
The readme says, this corresponds to the --global-work-size in official miner.
--cl-local-work Set the OpenCL local work size. Default is 64
--cl-global-work Set the OpenCL global work size as a multiple of the local work size. Default is 4096 * 64
Is -ethi 8 the same as these default settings?
what does -ethi 9 or 10 or 11, ... ?
The readme says, this corresponds to the --global-work-size in official miner.
--cl-local-work Set the OpenCL local work size. Default is 64
--cl-global-work Set the OpenCL global work size as a multiple of the local work size. Default is 4096 * 64
Is -ethi 8 the same as these default settings?
what does -ethi 9 or 10 or 11, ... ?
March 18, 2018, 03:33:53 AM
Bummer, but OK. I'll spit the cards up into two - AMD / NVidia and see what the pool reports.
OK, so as promised, I split the rig up into NVidia and AMD seperately to see what happens. Only NVidia submitted stale shares. I even went and ran the pairs as 1 x RX 470 and 2 x GTX 1050ti, each with a hashrate of ~28MH/s.
Here is the strange one, in this setup (2 x Claymore instances, NVidia + AMD), the rig submitted less stale shares as when compared to a single Claymore instance with the cards mixed.. But it was not really stable. NVidia kept on restarting.
I am back to the mixed setup where I have a 6 - 9% stale share rate for the past 12 Hours.
Some rig info for the 28 / 28 test:
1 x RX 470, 1500 mem straps, memory OC to 2000MHz.
2 x GTX 1050Ti, +650MHz memory OC, - 100MHz CPU UC, 75% Power
Latest Windows 10 Pro, Claymore excluded from Defender
Core2Duo E8400, 4GB Ram, 16GB Virtual memory
~17% CPU Usage, 46% memory usage whilst login with TeamViewer
Polly keeping an eye on things
March 18, 2018, 02:47:16 AM
If anyone has any insight on this, been reading/searching for some time and cannot figure it out:
Have a single rig with 4x1050 TI and 1xRX 560 (yes, I'm small time - but sorry, already taken, ladies). Running v11.4 as 11.5 keeps locking the PC up completely. AMD drivers are 17.12.1. Mining in ETH only mode. Zero hardware errors on HWInfo and no BIOS/clock/voltage/power modifications on the AMD.
The AMD card is showing a hashrate of ~9.18 vs ~13.5 on all the Nvidia cards. Miner showing 0 rejected/incorrect shares after 24 hours of mining. AMD card has right at 50% of the average shares of the 4 Nvidias - on nanopool if that matters. Looking at the submissions from the logs and all the shares from the Nvidia cards are accepted within ~170ms, but the AMD are ~2200ms?
Trying to understand the variance on the AMD - if I go by the hash rate vs the Nvidias, it should be closer to 68% not 50%, and what is the deal with the shares taking so long to be accepted from this one card? Thanks
Have a single rig with 4x1050 TI and 1xRX 560 (yes, I'm small time - but sorry, already taken, ladies). Running v11.4 as 11.5 keeps locking the PC up completely. AMD drivers are 17.12.1. Mining in ETH only mode. Zero hardware errors on HWInfo and no BIOS/clock/voltage/power modifications on the AMD.
The AMD card is showing a hashrate of ~9.18 vs ~13.5 on all the Nvidia cards. Miner showing 0 rejected/incorrect shares after 24 hours of mining. AMD card has right at 50% of the average shares of the 4 Nvidias - on nanopool if that matters. Looking at the submissions from the logs and all the shares from the Nvidia cards are accepted within ~170ms, but the AMD are ~2200ms?
Trying to understand the variance on the AMD - if I go by the hash rate vs the Nvidias, it should be closer to 68% not 50%, and what is the deal with the shares taking so long to be accepted from this one card? Thanks
March 17, 2018, 11:57:52 PM
Added claymore v11.5 to EthMonitoring.com
March 17, 2018, 11:33:14 PM
Do you mine running HwInfo 24/7 can affect hashrate/valid shares and overall work of the miner?
Yes, it can affect overall stability of a system
I have just recently started running HWinfo 24/7 and have not had any problems - so far - but that said it is best not to have anything else running when mining as it could affect the overall stability.
Yes, I've had the same with hwinfo. I installed and ran it and lost 5-10% hashrate. Turned it off, and back to normal.
March 17, 2018, 11:12:37 PM
Do you mine running HwInfo 24/7 can affect hashrate/valid shares and overall work of the miner?
Yes, it can affect overall stability of a system
I have just recently started running HWinfo 24/7 and have not had any problems - so far - but that said it is best not to have anything else running when mining as it could affect the overall stability.
March 17, 2018, 11:01:19 PM
Previously dcri was 30 by default in v10.6 - ETH only mining
Now in 11.5 if it is not specified claymore auto-detects the best dcri for the gpu
With the auto-detect I get dcri values 14,20,8 and the system is not stable and crashes ever few hours despite using -oldkernels 1
whereas with the 10.6 which had default dcri of 30.. never had crashes
Now in 11.5 if it is not specified claymore auto-detects the best dcri for the gpu
With the auto-detect I get dcri values 14,20,8 and the system is not stable and crashes ever few hours despite using -oldkernels 1
whereas with the 10.6 which had default dcri of 30.. never had crashes
March 17, 2018, 09:54:48 PM
People refuse to read the readme file supplied with each miner.
So lazy. Want instant answers without putting any effort into it.
So lazy. Want instant answers without putting any effort into it.
janding - grow up. I've read the readme several times and this thread. I still can't figure out how to get it work. The readme does not mention a subfolder, just filename. I'm running windows.
Honestly, try another miner. A lot of better ones out there now ... with more reasonable dev fees
Jump to: