3.5b same result, log:
1970.01.01:07:15:48.496: unkn OpenCL driver version: 18.5.1
...
I tried the 3.5b. No change:
miner freeze
here log
...
Thank you for reporting this problem. We have send a PM to each of you with additional information and questions that will help us resolve it.
-clgreen 1 what is this parameter
Activates new power-efficient kernels for AMD RX460/470/480/560/570/580. They aren't faster than the normal kernels (in most cases are slightly slower but not always) but should lower your power consumption by 2-3 W per card (RX570). Whether they are more optimal for you depends on the price of electricity and the settings of your cards. Please note that you have to run auto-tune again to find the optimal GT settings for these kernels as they are different from the normal kernels.
Good news!
Almost every message after releasing new version is about problems (which is quite normal and alright), so it's quite nice to get a few positive reports. Don't hesitate to write us with any ideas for improvements, etc.
Did you see:
https://bitcointalksearch.org/topic/m.43145022 Yes, thank you for this suggestion. We are planning much more feature-rich remote management API in the future releases (we will keep compatibility with Claymore's as fallback). Note that you can get the per-GPU hashrate with the current API. The Nvidia overclocking options are planned for the next release.
PhoenixMiner 3.5b works stable for my only 2x Sapphire RX580 Nitro+ rig but I don't see any change on my GPUs power draw when using new green kernels.
I get same hashrate (30.3MH) with same GPU only power draw in GPU-Z sensors tab (90W for every card) when using new kernels.
You should see about 2-3W per card but probably not with GPU-Z. We have measured 2.4W per RX570 with settings 1100/2000 MHz and 870mV core voltage. The hashrate drop was from 29.15 to 29.0 MH/s.
green kernel 1275w 304mh/s, without 308mh/s 1287w on wattmeter.
Please make sure that you have auto-tuned the green kernels (the old GT values are not optimal for the green kernels). Sometimes it may take same time before you see the real difference in power (30 mintues or so after starting). Also the drops in the power and in the hashrate depend on the settings of the card and the ratio between the core clock and memory clocks, as well as the memory timings. The green kernels can be used only on some gpus by specifying -clgrean for each gpu individually: for example -clgreen 1,1,0,0,0 will run the green kernels only on the first two GPUs in five GPU rig.
miner freeze v3.5b detects all devices but then ends with error !!!!
please HELP
Can you tell us if you see any error message window, or the miner just freezes without showing anything?
Здpaвcтвyйтe кoллeги мaйнepы!
cтoлкнyлcя c пpoблeмoй пpи нacтpoйкe мaйнepa PhoenixMiner , в OS HiveOS
Hacтpaивaл coглacнo oпиcaния в пocтy:
https://forum.hiveos.farm/t/custom-miner-integration/4521/212?u=chivekubikнacтpoить yдaлocь, нo пocлe зaпycкa мaйнepa, cмoтpю в лoги и вижy:
Eth: New job #2699da4d from eu1-etc.ethermine.org:4444; diff: 4000MH
Eth speed: 0.000 MH/s, shares: 0/0/0, time: 0:01
GPUs: 1: 0.000 MH/s (0) 2: 0.000 MH/s (0) 3: 0.000 MH/s (0)
Eth speed: 0.000 MH/s, shares: 0/0/0, time: 0:01
GPUs: 1: 0.000 MH/s (0) 2: 0.000 MH/s (0) 3: 0.000 MH/s (0)
Eth speed: 0.000 MH/s, shares: 0/0/0, time: 0:01
GPUs: 1: 0.000 MH/s (0) 2: 0.000 MH/s (0) 3: 0.000 MH/s (0)
Eth speed: 0.000 MH/s, shares: 0/0/0, time: 0:01
GPUs: 1: 0.000 MH/s (0) 2: 0.000 MH/s (0) 3: 0.000 MH/s (0)
Eth: New job #8e242334 from eu1-etc.ethermine.org:4444; diff: 4000MH
xeшpeйт нe пoднимaeтcя в тeчeниe 15 мин.
Teмпepaтypa GPU нe пoднимaeтcя
________
Moe жeлeзo:
OS:
DISTRIB_ID=Ubuntu
DISTRIB_RELEASE=16.04
DISTRIB_CODENAME=xenial
DISTRIB_DESCRIPTION=“Ubuntu 16.04.4 LTS”
__
GPU:
GeForce GTX 1080 8119 MiB · MSI - 3 шт.
____
Пpилaгaю cкpиншoты c нacтpoйкaми
https://prnt.sc/ksozceи xeшpeйтoм
https://prnt.sc/ksozsiФepмa
https://prnt.sc/kspa8p_______________
Пoмoгитe пoжaлyйcтa, нacтpoить кoppeктнo мaйнep
Can you please send us (or post here) the first few pages from the log file? It seems that the DAG generation never finishes but we can't be sure because we can see the start of the log in your screenshot, and then an excerpt after a few minutes but not what happens in between.
BUG report: Linux version PM 3.5a has support for fan/temp only upto 7 devices.
http://prntscr.com/kspvgcTemperature monitoring will not work on 7+ devices, otherwise it seems the fastest and most efficient ethash miner to me. Good job!
EDIT: Does rmode on linux execs reboot.bat or reboot.bash/sh ?
EDIT2: It does read temp/fan till PCI 0x09, miner marks it as 0x10 but its actually 0x0a.
Thank you for reporting this problem. We haven't tested the Linux version on a rig with more than 6 GPUs yet and sure enough, there is a problem. We have found the issue and it will be fixed in 3.5b, which will be released later today or first thing tomorrow after we are sure that there aren't any more problems that we can find here.
Hi all, just reporting my results... on:
PhoenixMiner 3.5a -- First release with support for Linux
4.10.17-hiveos #1 SMP Sun Dec 10 08:48:43 EET 2017 x86_64 x86_64 x86_64 GNU/Linux
DISTRIB_ID=Ubuntu
DISTRIB_RELEASE=16.04
DISTRIB_CODENAME=xenial
DISTRIB_DESCRIPTION="Ubuntu 16.04.5 LTS"
Based on this post over at HiveOS community:
https://forum.hiveos.farm/t/custom-miner-integration/4521/2122-3 MH/s increase (reported hashrate) over Claymore 11.9,
BUT a significant increase in stale shares! [using
https://www.hiveon.net/ ETH pool]
screenshot:
Thank you, this is very helpful. Most probably the miner can't recognize the OpenCL driver and some of the most aggressive driver-specific optimizations aren't used. Can you please check if you have a line like this:
2018.09.10:06:24:11.757: main OpenCL driver version: 17.40
in the start of PhoenixMiner log (it should be among the first five to ten lines in the log file)? Also, could you please tell us the version of your AMD drivers?
