Author

Topic: [ANN] dstm's ZCash / Equihash Nvidia Miner v0.6.2 (Linux / Windows) - page 110. (Read 224961 times)

newbie
Activity: 17
Merit: 0

Sound like a driver crash, odd that the miner doesn't crash either, you could try to kill all binaries, and do a nvidia drivers reset, then restart the miner, may work, your solution work too Smiley are you using latest version with the --noconnect ?

Me i noticed another odd thing, if you use the miner multiple times throught the day, the hashrate lower gradually like 1% each time it's started, if you reboot the rig it's back up to the normal hashrate, i never noticed this with EWBF, it's kinda odd.

Yep, latest zm with --noconnect, latest nvidia and cuda driver.

For now ewbf's still a win for me, despite a little slower than zm, but I don't have to worry either pools going down, or have to restart the miner every X hours. I like zm with active dev and hope with constant update, the stability and features is up there or better then with ewbf.
I'm, still requesting pool failover, by the way  Wink

I'm getting 390 vs 375 with my 1070 (60%, +175 core, +200 Mem)
But with dev fee the added 15 sol/s turns into a net +7.5 sol
Would be nice to have the dev fee to drop to 1% Smiley
Great work

I've got 478-489 sol/s Gigabyte Aorus GTX 1070 with micron memory, 70% power, +100 core, +700 memory

Has anyone else noticed that when you drop the power draw on your cards, it doesn't adjust the Sol/W readings in the miner? I took mine from 85% down to 67% and after a few minutes they were still showing 3.30 Sol/W.

Seems like driver bug? I noticed after add more card, Sol/W reading shows "infinity" in my rig. I have to reinstall nvidia and cuda driver then the Sol/W reading back to normal.
newbie
Activity: 11
Merit: 0
Has anyone else noticed that when you drop the power draw on your cards, it doesn't adjust the Sol/W readings in the miner? I took mine from 85% down to 67% and after a few minutes they were still showing 3.30 Sol/W.
newbie
Activity: 44
Merit: 0
DTSM is actively developing his program so I have no issues paying a dev fee to support him. This dev fee may be meaningless if he significantly boosts the hash rate with future builds.
jr. member
Activity: 557
Merit: 5
I'm getting 390 vs 375 with my 1070 (60%, +175 core, +200 Mem)
But with dev fee the added 15 sol/s turns into a net +7.5 sol
Would be nice to have the dev fee to drop to 1% Smiley
Great work
hero member
Activity: 756
Merit: 507
after long tests and usage my opinion about dstm for GTX1080ti is here - it is the same as ewbf
why? here is why :
hashrate for one gtx 1080ti aprox  770sol, so 4 x 760 = 3040 sol/s
while with same settings it is 2960 sol on ewbf.
so you it is 2.7% win for dstm.  but, there is a dev fee 2% and ewbf we can use without devfee - that means that actual difference is less then 1%
and taking into account, that ewbf is more stable for my rig, the choice is obvious

however! for 2 x gtx1070 on dstm i'm getting 960 sol/s and 920 sol/s with ewbf 
so it is 4.3% difference, minus 2% dev fee and we comes to 2.3% benefit.

I can accept that bonus for gtx1070
full member
Activity: 224
Merit: 100
CryptoLearner

You can probably get something similar with setting up a batch script that after it exit the miner because of disconnect (use the --noreconnect miner switch) it restart the miner on a different pool. Have to be sure the miner exit on pool error. But should work


Hey that's a good idea, but I don't think I can do that in my rig,

I've got a strange experience using zm after running 15 hours give or take, I received an e-mail from flypool, said that my miner was offline. Then I checked my rig, Internet connection ok, zm still running and mining but not sending any shares to the pool (only reporting my hashrate, temp, etc) therefore the pool sent an e-mail.

I tried to close the miner and restart it again, everything going back to normal. Now I have to set my batch file to auto restart zm every 12 hours, been running good for about a week. I tried to make a loop like you suggested but it didn't work, the miner didn't auto restarted like it used to be, it just opened new cmd window with the new backup pool address but not mining because the gpus somewhat still minning in the first cmd.  Huh

Sound like a driver crash, odd that the miner doesn't crash either, you could try to kill all binaries, and do a nvidia drivers reset, then restart the miner, may work, your solution work too Smiley are you using latest version with the --noconnect ?

Me i noticed another odd thing, if you use the miner multiple times throught the day, the hashrate lower gradually like 1% each time it's started, if you reboot the rig it's back up to the normal hashrate, i never noticed this with EWBF, it's kinda odd.
newbie
Activity: 17
Merit: 0

You can probably get something similar with setting up a batch script that after it exit the miner because of disconnect (use the --noreconnect miner switch) it restart the miner on a different pool. Have to be sure the miner exit on pool error. But should work


Hey that's a good idea, but I don't think I can do that in my rig,

I've got a strange experience using zm after running 15 hours give or take, I received an e-mail from flypool, said that my miner was offline. Then I checked my rig, Internet connection ok, zm still running and mining but not sending any shares to the pool (only reporting my hashrate, temp, etc) therefore the pool sent an e-mail.

I tried to close the miner and restart it again, everything going back to normal. Now I have to set my batch file to auto restart zm every 12 hours, been running good for about a week. I tried to make a loop like you suggested but it didn't work, the miner didn't auto restarted like it used to be, it just opened new cmd window with the new backup pool address but not mining because the gpus somewhat still minning in the first cmd.  Huh
full member
Activity: 224
Merit: 100
CryptoLearner
Hello dstm, may I suggest add failover pools feature like in ewbf and cgminer, thanks.

I second this feature request.

(I actually thought it was already in there and I just didnt get around to setting it up yet)

You can probably get something similar with setting up a batch script that after it exit the miner because of disconnect (use the --noreconnect miner switch) it restart the miner on a different pool. Have to be sure the miner exit on pool error. But should work

kinda like

Code:
::looppool
your zm cmline pool 1 --noreconnect
goto pool 2

:: pool 2
your zm cmline pool 2 --noreconnect
goto pool 3

:: pool 3
your zm cmline pool 3 --noreconnect
goto looppool

but i agree, config files and pool failover are awesome Wink
newbie
Activity: 44
Merit: 0
Hello dstm, may I suggest add failover pools feature like in ewbf and cgminer, thanks.

I second this feature request.

(I actually thought it was already in there and I just didnt get around to setting it up yet)
newbie
Activity: 17
Merit: 0
Hello dstm, may I suggest add failover pools feature like in ewbf and cgminer, thanks.
newbie
Activity: 126
Merit: 0
Я ничeгo нe cocтaвлял. Bce кaк былo изнaчaльнo, тoлькo зaмeнил кoшeлeк.
Boт вce чтo в бaтникe.

Code:
echo Don't forget to change your pool and login information.

zm --server eu1-zcash.flypool.org  --port 3333 --user t1NFooZ82fzQeWWUH27G8uxLJvNfwMnrBKG


@echo off
REM Usage:
REM  zm --server hostname --port port_nr --user user_name
REM     [--pass password] [options]...
REM
REM  Stratum:
REM     --server         Stratum server hostname
REM                      prefix hostname with 'ssl://' for encrypted
REM                      connections - e.g. ssl://mypool.com
REM     --port           Stratum server port number
REM     --user           Username / Workername
REM     --pass           Worker password
REM
REM  Options:
REM     --help           Print this help
REM     --dev            Space separated list of cuda devices
REM     --list-devices   List available cuda devices
REM     --time           Enable output of timestamps
REM     --logfile        [=path] Append logs to the file named by 'path'
REM                      If 'path' is not given append to 'zm.log' in
REM                      current working directory.
REM     --noreconnect    Disable automatic reconnection on network errors.
REM
REM     --temp-target    In C - If set, enables temperature controller.
REM                      The workload of each GPU will be continuously
REM                      adjusted such that the temperature stays around
REM                      this value. It is recommended to set your fan speed
REM                      to 100% when using this setting.
REM
REM     --telemetry      [=ip:port]. Starts telemetry server. Telemetry data
REM                      can be accessed using a web browser(http) or by json-rpc.
REM                      If no arguments are given the server listens on
REM                      127.0.0.1:2222 - Example: --telemetry=0.0.0.0:2222
REM                      Valid port range [1025-65535]
REM  Example:
REM     zm --server servername.com --port 1234 --user username



Я вижy твoю cтaтиcтикy, вce нopм
newbie
Activity: 44
Merit: 0
EthMonitoring.com EthControl software has been updated to ZM dstm 0.5.5 version.

Nokia could you add luckpool.org to ethmonitoring.com?
full member
Activity: 224
Merit: 100
CryptoLearner
Been running v0.5.4 on 4 rigs, all 1080ti for 2 days, zero issues. Just changed them all over to this latest version v0.5.5 (using Tight VNC, which is totally free, to remotely roll it out to all my rigs)

What I've Noticed
1. There is a definite improvement in hash rate as indicated by my miner stats in my pool (luckpool.org).
2. The GPU's run MUCH cooler than they did with EWBF. So much cooler that I had to confirm that ZM was actually working as the fans on some cards were not even spinning some times.

Has any one else noticed that their GPU's run cooler?

Overall all big thanks for this version as there is an increase in my profit and it has been rock solid for me for days.

Feature Request
Some colour would be nice like EWBF has.
This way you can tell at a glance if, for example, your rigs are over heating (ie a fan dies, this actually happened to me yesterday)
Being able to choose what stats show what colours would be awesome.

what is the improvement in hashrate ? how much do you think 0.5.5 improves over EWBF, keep in mind also you can turn off the dev fee on EWBF while here it's hard coded at 2%... Would be great if you can provide some evidence for your claims...

For me the improvement is about 6-7% (2680sols with EWBF and 2850sols with ZM), counting the 2% fee that's about 4-5% so not too bad, my cards run the same between 0.5.4 and 0.5.5, no improvement of hashrate or temperatures, but it work pretty well and pretty stable, so good job overall imho. Haven't had any issues with disconnection or anything.
legendary
Activity: 2128
Merit: 1657
Been running v0.5.4 on 4 rigs, all 1080ti for 2 days, zero issues. Just changed them all over to this latest version v0.5.5 (using Tight VNC, which is totally free, to remotely roll it out to all my rigs)

What I've Noticed
1. There is a definite improvement in hash rate as indicated by my miner stats in my pool (luckpool.org).
2. The GPU's run MUCH cooler than they did with EWBF. So much cooler that I had to confirm that ZM was actually working as the fans on some cards were not even spinning some times.

Has any one else noticed that their GPU's run cooler?

Overall all big thanks for this version as there is an increase in my profit and it has been rock solid for me for days.

Feature Request
Some colour would be nice like EWBF has.
This way you can tell at a glance if, for example, your rigs are over heating (ie a fan dies, this actually happened to me yesterday)
Being able to choose what stats show what colours would be awesome.

what is the improvement in hashrate ? how much do you think 0.5.5 improves over EWBF, keep in mind also you can turn off the dev fee on EWBF while here it's hard coded at 2%... Would be great if you can provide some evidence for your claims...
newbie
Activity: 44
Merit: 0
Been running v0.5.4 on 4 rigs, all 1080ti for 2 days, zero issues. Just changed them all over to this latest version v0.5.5 (using Tight VNC, which is totally free, to remotely roll it out to all my rigs)

What I've Noticed
1. There is a definite improvement in hash rate as indicated by my miner stats in my pool (luckpool.org).
2. The GPU's run MUCH cooler than they did with EWBF. So much cooler that I had to confirm that ZM was actually working as the fans on some cards were not even spinning some times.

Has any one else noticed that their GPU's run cooler?

Overall all big thanks for this version as there is an increase in my profit and it has been rock solid for me for days.

Feature Request
Some colour would be nice like EWBF has.
This way you can tell at a glance if, for example, your rigs are over heating (ie a fan dies, this actually happened to me yesterday)
Being able to choose what stats show what colours would be awesome.
sr. member
Activity: 392
Merit: 266
EthMonitoring.com
EthMonitoring.com EthControl software has been updated to ZM dstm 0.5.5 version.
full member
Activity: 224
Merit: 100
CryptoLearner
Bceм пpивeт. Ктo мoжeт пoдcкaзaть пoчeмy нe нaчиcляютcя zec зa мaйнинг? Oнo зaчиcляeтcя oт кaкoй тo oпpeдeлeннoй cyммы или пo вpeмeни? Cтoялo нoчь, peзyльтaт пo 0. Bидeoкapтa gtx 1050 2gb. B paзгoнe выдaeт 164 sol. Кoшeлeк Zec пpoпиcaл в start.bat

Знaчит нe пpaвильнo бaтник cocтaвил.

I have this error

2017-11-16 11:54:22|gpu_id 1 2 000000052AB60400 unknown error
2017-11-16 11:54:22|gpu_id 3 2 000000055F600000 unknown error
2017-11-16 11:54:22|gpu_id 4 0 0000000579C00000 unknown error
2017-11-16 11:54:22|gpu_id 0 0 000000050DEB0200 unknown error
2017-11-16 11:54:22|gpu_id 2 6 00000005450420C0 unknown error
2017-11-16 11:54:22|gpu 2 unresponsive - check overclocking
2017-11-16 11:54:22|cudaMemcpy 1 failed
2017-11-16 11:54:22|gpu 4 unresponsive - check overclocking
2017-11-16 11:54:22|cudaMemcpy 1 failed
2017-11-16 11:54:22|gpu 1 unresponsive - check overclocking
2017-11-16 11:54:22|cudaMemcpy 1 failed
2017-11-16 11:54:22|gpu 0 unresponsive - check overclocking
2017-11-16 11:54:22|gpu 3 unresponsive - check overclocking
2017-11-16 11:54:22|cudaMemcpy 1 failed
2017-11-16 11:54:22|cudaMemcpy 1 failed

Hello, i got 2x10rigs (all 1070), and your miner work awesome on the first 10 rigs (+7% hash vs EWBF), The other 10 rigs have a different GPU brand and same hashpower (all 1070) but i can't get your miner to work, i experience the exact same error as bruce81, even with base clock/power on cards, i also get "memory errors" log line, but thoses rigs have worked overclocked flawlessly for months on EWBF (and still does) (All 20rigs have the same gpu, drivers, OS, SSD, ram, and so on, only the brand of the GPU change (10x rigs with EVGA's 1070, 10x rigs with KFA2 1070's)

An idea what it could be ? Im willing to do some testing or advanced logging.

I was wondering if you could also add something to tweak the intensity? Maybe i need to lower it a bit so that the miner doesn't explode, dunno Smiley

Regards.
newbie
Activity: 2
Merit: 0
Я ничeгo нe cocтaвлял. Bce кaк былo изнaчaльнo, тoлькo зaмeнил кoшeлeк.
Boт вce чтo в бaтникe.

Code:
echo Don't forget to change your pool and login information.

zm --server eu1-zcash.flypool.org  --port 3333 --user t1NFooZ82fzQeWWUH27G8uxLJvNfwMnrBKG


@echo off
REM Usage:
REM  zm --server hostname --port port_nr --user user_name
REM     [--pass password] [options]...
REM
REM  Stratum:
REM     --server         Stratum server hostname
REM                      prefix hostname with 'ssl://' for encrypted
REM                      connections - e.g. ssl://mypool.com
REM     --port           Stratum server port number
REM     --user           Username / Workername
REM     --pass           Worker password
REM
REM  Options:
REM     --help           Print this help
REM     --dev            Space separated list of cuda devices
REM     --list-devices   List available cuda devices
REM     --time           Enable output of timestamps
REM     --logfile        [=path] Append logs to the file named by 'path'
REM                      If 'path' is not given append to 'zm.log' in
REM                      current working directory.
REM     --noreconnect    Disable automatic reconnection on network errors.
REM
REM     --temp-target    In C - If set, enables temperature controller.
REM                      The workload of each GPU will be continuously
REM                      adjusted such that the temperature stays around
REM                      this value. It is recommended to set your fan speed
REM                      to 100% when using this setting.
REM
REM     --telemetry      [=ip:port]. Starts telemetry server. Telemetry data
REM                      can be accessed using a web browser(http) or by json-rpc.
REM                      If no arguments are given the server listens on
REM                      127.0.0.1:2222 - Example: --telemetry=0.0.0.0:2222
REM                      Valid port range [1025-65535]
REM  Example:
REM     zm --server servername.com --port 1234 --user username
newbie
Activity: 126
Merit: 0
Bceм пpивeт. Ктo мoжeт пoдcкaзaть пoчeмy нe нaчиcляютcя zec зa мaйнинг? Oнo зaчиcляeтcя oт кaкoй тo oпpeдeлeннoй cyммы или пo вpeмeни? Cтoялo нoчь, peзyльтaт пo 0. Bидeoкapтa gtx 1050 2gb. B paзгoнe выдaeт 164 sol. Кoшeлeк Zec пpoпиcaл в start.bat

Знaчит нe пpaвильнo бaтник cocтaвил.

I have this error

2017-11-16 11:54:22|gpu_id 1 2 000000052AB60400 unknown error
2017-11-16 11:54:22|gpu_id 3 2 000000055F600000 unknown error
2017-11-16 11:54:22|gpu_id 4 0 0000000579C00000 unknown error
2017-11-16 11:54:22|gpu_id 0 0 000000050DEB0200 unknown error
2017-11-16 11:54:22|gpu_id 2 6 00000005450420C0 unknown error
2017-11-16 11:54:22|gpu 2 unresponsive - check overclocking
2017-11-16 11:54:22|cudaMemcpy 1 failed
2017-11-16 11:54:22|gpu 4 unresponsive - check overclocking
2017-11-16 11:54:22|cudaMemcpy 1 failed
2017-11-16 11:54:22|gpu 1 unresponsive - check overclocking
2017-11-16 11:54:22|cudaMemcpy 1 failed
2017-11-16 11:54:22|gpu 0 unresponsive - check overclocking
2017-11-16 11:54:22|gpu 3 unresponsive - check overclocking
2017-11-16 11:54:22|cudaMemcpy 1 failed
2017-11-16 11:54:22|cudaMemcpy 1 failed
newbie
Activity: 2
Merit: 0
Bceм пpивeт. Ктo мoжeт пoдcкaзaть пoчeмy нe нaчиcляютcя zec зa мaйнинг? Oнo зaчиcляeтcя oт кaкoй тo oпpeдeлeннoй cyммы или пo вpeмeни? Cтoялo нoчь, peзyльтaт пo 0. Bидeoкapтa gtx 1050 2gb. B paзгoнe выдaeт 164 sol. Кoшeлeк Zec пpoпиcaл в start.bat
Jump to: