PhoenixMiner 6.2c: fastest Ethereum/Ethash miner with lowest devfee (Win/Linux)

[crypper]

i'm afraid of heat 

I mean, why don't you set the fan speed to fixed 90/100%?
it won't hurt anything and surely will help to cool memory and VRM chips.

surely it will hurt fans' bearings

[j2james]

Hi PM developers.

Today I have updated PM from 3.0c to 3.5d. And I cannot connect to CDM remote monitoring. I did not change -cdmport and -cdmpass.

I use command line parameters (not config file): -cdm 2 -cdmport 3333 -cdmpass xxxxx

Any ideas?

I have checked with config file. The same problem.

After back to 3.0c all works (with command line parameters and with config file)

[happyleper]

I always use custom fan curve setting in MSI Afterburner or Sapphire Trixx. I set the fan speed to 5 points above the GPU's temperature. For example, if the GPU's temp is 52C, then the fan will spin at speed of 57%, 65C temp 70% fan, and it always spin based on current GPU's temp. Can this custom fan curve setting be done via PhoenixMiner?

ALWAYS use 3rd party soft to control clocks, power and fans: OverdriveNtool for red and nvidiaInspector for green; don't use AB - it's shit. Don't use miners for controls.

AB works just fine. And will apply clocks at boot. Can nvidia inspector do that?

With a properly configured batch file that runs at startup, yes, it can.

[gsrcrxsi314]

I always use custom fan curve setting in MSI Afterburner or Sapphire Trixx. I set the fan speed to 5 points above the GPU's temperature. For example, if the GPU's temp is 52C, then the fan will spin at speed of 57%, 65C temp 70% fan, and it always spin based on current GPU's temp. Can this custom fan curve setting be done via PhoenixMiner?

ALWAYS use 3rd party soft to control clocks, power and fans: OverdriveNtool for red and nvidiaInspector for green; don't use AB - it's shit. Don't use miners for controls.

AB works just fine. And will apply clocks at boot. Can nvidia inspector do that?

With a properly configured batch file that runs at startup, yes, it can.

So natively, it cant. Thanks.

[lemonyo30]

I need Help please, im getting this error below, changed as well 5555 still the same error

No CUDA driver found
Available GPUs for mining:
GPU1: Radeon RX 580 Series (pcie 2), OpenCL 2.0, 8 GB VRAM, 36 CUs
GPU2: Radeon RX 580 Series (pcie 3), OpenCL 2.0, 8 GB VRAM, 36 CUs
Listening for CDM remote manager at port 3333 in full mode
Eth: the pool list contains 4 pools
Eth: primary pool: eu1.ethermine.org:4444
Starting GPU mining
Eth: Connecting to ethash pool eu1.ethermine.org:4444 (proto: QtMiner)
GPU1: 24C 70%, GPU2: 21C 70%
Eth: Could not connect to ethash pool eu1.ethermine.org:4444: A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond
Eth: Reconnecting in 20 seconds...

[styleshifter]

Hey phoenix miner developers.
You should contact all pools like ethermine… because they need to update their websites that Phoenix Miner for Linux is now available!

[happyleper]

I always use custom fan curve setting in MSI Afterburner or Sapphire Trixx. I set the fan speed to 5 points above the GPU's temperature. For example, if the GPU's temp is 52C, then the fan will spin at speed of 57%, 65C temp 70% fan, and it always spin based on current GPU's temp. Can this custom fan curve setting be done via PhoenixMiner?

ALWAYS use 3rd party soft to control clocks, power and fans: OverdriveNtool for red and nvidiaInspector for green; don't use AB - it's shit. Don't use miners for controls.

AB works just fine. And will apply clocks at boot. Can nvidia inspector do that?

With a properly configured batch file that runs at startup, yes, it can.

So natively, it cant. Thanks.

With a power reduction on a 9 GPU rig of nearly 20 percent using NVIDIA Inspector over Afterburner, I think running a batch file at startup is a small price to pay.

https://www.reddit.com/r/EtherMining/comments/9ah7qc/how_to_tune_nvidia_gpu_for_best_efficiency_nvidia/

[gsrcrxsi314]

I always use custom fan curve setting in MSI Afterburner or Sapphire Trixx. I set the fan speed to 5 points above the GPU's temperature. For example, if the GPU's temp is 52C, then the fan will spin at speed of 57%, 65C temp 70% fan, and it always spin based on current GPU's temp. Can this custom fan curve setting be done via PhoenixMiner?

ALWAYS use 3rd party soft to control clocks, power and fans: OverdriveNtool for red and nvidiaInspector for green; don't use AB - it's shit. Don't use miners for controls.

AB works just fine. And will apply clocks at boot. Can nvidia inspector do that?

With a properly configured batch file that runs at startup, yes, it can.

So natively, it cant. Thanks.

With a power reduction on a 9 GPU rig of nearly 20 percent using NVIDIA Inspector over Afterburner, I think running a batch file at startup is a small price to pay.

https://www.reddit.com/r/EtherMining/comments/9ah7qc/how_to_tune_nvidia_gpu_for_best_efficiency_nvidia/

sounds like he didnt know how to use AB to its potential. if you leave the power limit high, and dont play with the clocks, then of course it'll have worse power consumption.

all you have to do is drag the power limit lower and BOOM, lower power draw. its not complicated. he had it at 70, but more or less replicated a lower power limit with a more complicated process using nv inspector.

personally i run mine at 65, and actually increase the core clock to keep hashrates the same. power cannot exceed the limit, so it forces lower voltage, essentially the same thing he did, just using a different tool.

theres nothing wrong with AB. you can achieve the same results with both. you just have to know how to use it properly.

[PhoenixMiner]

I am trying out this miner with good results so far.

I do however have one problem that I cannot seem to resolve. No matter how many pools I enter into epools.txt the miner ALWAYS reports that I have one more pool then I have actually entered. If I only have a single line in epools.txt the miner reports two pools when it starts, it tries to connect to the first one and fails because it simply does not exist ("could not resolve host"). After some time it tries the second pool (witch in fact is the first and only one) and it starts mining.


This example is working perfectly fine with claymore, but not with phoenix.

Code:

POOL: daggerhashimoto.eu.nicehash.com:3353, WALLET: <mywallet.workename>, PSW: x, ESM: 3, ALLPOOLS: 1

I have the same problem on quite a few rigs. They are all running HiveOS (latest version).

Any suggested solutions?

  Please note that the first one or two pools from the pool list may come from the command line. If you have specified one of the options -pool, -wal, -pass, -coin, or -proto on the command line (perhaps via the custom miner settings of the HiveOS control panel), this will specify a pool and since it is the first pool in the list, PhoenixMiner will first try to connect to it. You just need to remove all these options from the command line and then only the pools in epools.txt will matter.

Code:

...

As you can see, it takes about 40 seconds to initialize the first 3 cards, the log is pretty much the same through the rest of the DAG creation.  It was able to get a valid share while the DAG was being created for the other cards so it's working as soon as the card is finished initializing.

on -gser 1 it would just list the first 3 cards:

GPU1 GPU1: Starting up... (0)
GPU2 GPU2: Starting up... (0)
GPU3 GPU3: Starting up... (0)

but it would never show "GPU1 Eth: Generating light cache for epoch #214" for any cards after that.

   Thank you for the log! Apparently, the light cache generation is not causing the problem but perhaps some incompatibility of the Linux kernels with some Nvidia cards (the light cache is created only once by whatever GPU starts first and then is used by all the other GPUs). We have included same measures in the development version of 3.6 to detect better what happens right after the light cache generation, and we will update the CUDA kernels for Linux, so hopefully this will solve the problem in the next version.

I always use custom fan curve setting in MSI Afterburner or Sapphire Trixx. I set the fan speed to 5 points above the GPU's temperature. For example, if the GPU's temp is 52C, then the fan will spin at speed of 57%, 65C temp 70% fan, and it always spin based on current GPU's temp. Can this custom fan curve setting be done via PhoenixMiner?

   No, with PhoenixMiner you can either fix the temperature (e.g. -tt 60) or the fan speed (with negative number after -tt). You can also set a maximum fan speed but that's it. We will update the hardware control options in Mil Shield 3.6 but such custom fan curves are beyond what we have at mind for now.
 

Hi PM developers.

Today I have updated PM from 3.0c to 3.5d. And I cannot connect to CDM remote monitoring. I did not change -cdmport and -cdmpass.

I use command line parameters (not config file): -cdm 2 -cdmport 3333 -cdmpass xxxxx

Any ideas?

I have checked with config file. The same problem.

After back to 3.0c all works (with command line parameters and with config file)

   The only change with respect to remote monitoring in 3.5d is that now it starts 10-20 seconds after the miner starts but otherwise it should be working fine - we are using it 24/7 on our test rigs without any problems. From our own experience, this problem is perhaps related to Windows Firewall, which blocks the CDM port for the executable of the new version (it should show you a pop-up window, asking if you want to open the port the first time you start the a new version of PhoenixMiner). Another possible reason (again related to Windows Firewall) is that it may "decide" that the local network is "public" instead of "private" and it has much more strict rules for public networks, so it again blocks the CDM port of PhoenixMiner. A quick test to see if the Firewall is the culprit is to turn it off for a short time and see if the remote management works.

I need Help please, im getting this error below, changed as well 5555 still the same error

Code:

No CUDA driver found
Available GPUs for mining:
GPU1: Radeon RX 580 Series (pcie 2), OpenCL 2.0, 8 GB VRAM, 36 CUs
GPU2: Radeon RX 580 Series (pcie 3), OpenCL 2.0, 8 GB VRAM, 36 CUs
Listening for CDM remote manager at port 3333 in full mode
Eth: the pool list contains 4 pools
Eth: primary pool: eu1.ethermine.org:4444
Starting GPU mining
Eth: Connecting to ethash pool eu1.ethermine.org:4444 (proto: QtMiner)
GPU1: 24C 70%, GPU2: 21C 70%
Eth: Could not connect to ethash pool eu1.ethermine.org:4444: A connection attempt failed because the connected party did not properly respond after a period of time, or established connection failed because connected host has failed to respond
Eth: Reconnecting in 20 seconds...

   There are two possible reasons for this: you either have connection problems -  check your network settings, DNS settings, etc., or you have left YourWallet as the wallet name in the command-line settings (the .bat file), or in the epools.txt file - in such case replace YourWallet with your actual wallet address.

[btckesh]

phoenixminer for linux, nice guys. this is what i've been waiting for and now i've almost overseen it! any differences compared to windows with 1070ti/1080 cards?

[Matkurb]

I always use custom fan curve setting in MSI Afterburner or Sapphire Trixx. I set the fan speed to 5 points above the GPU's temperature. For example, if the GPU's temp is 52C, then the fan will spin at speed of 57%, 65C temp 70% fan, and it always spin based on current GPU's temp. Can this custom fan curve setting be done via PhoenixMiner?

ALWAYS use 3rd party soft to control clocks, power and fans: OverdriveNtool for red and nvidiaInspector for green; don't use AB - it's shit. Don't use miners for controls.

AB works just fine. And will apply clocks at boot. Can nvidia inspector do that?

With a properly configured batch file that runs at startup, yes, it can.

So natively, it cant. Thanks.

With a power reduction on a 9 GPU rig of nearly 20 percent using NVIDIA Inspector over Afterburner, I think running a batch file at startup is a small price to pay.

https://www.reddit.com/r/EtherMining/comments/9ah7qc/how_to_tune_nvidia_gpu_for_best_efficiency_nvidia/

That is great. I will give it a try.

[AZuley]

The same question. Does not work on 1060 3Gb Ubuntu 16-18

https://preview.ibb.co/iw8X9K/phoenix.jpg

Kernel: 4.13.16
Drivers: N 396.54

   Thank you for the system information. We have successfully ran 1060 3GB and 6GB under HiveOS 0.5-76 so we are really not sure why is this happening. Probably it is somehow connected to the number of the GPUs as all reports so far are with a lot of GPUs. Please try to add -gpus 123 command-line parameter in your custom miner settings in order to use only the first three GPUs and see if this fixes the problem. Obviously this isn't a workable solution but it will help us reproduce it.

  I tried on your advice - nothing has changed with 3 cards. The result is the same.

[Silverhydra57]

Hi,
Sorry if the problem has already been solved, I did not read the entire thread.

I am willing to give this miner a try but it draws far more power than claymore for a similar or little higher hashrate.

I use it with win10 with 6 RX580.
With claymore (11.9), 186Mh/s @ 830W
With Phoenix (3.5d), 187Mh/s @ 1150W!
I tried to use the same values for clocks and voltages (which are set in config.txt) or to let it blank without any effect.
What am I doing wrong?

[crypper]

Hi,
Sorry if the problem has already been solved, I did not read the entire thread.

I am willing to give this miner a try but it draws far more power than claymore for a similar or little higher hashrate.

I use it with win10 with 6 RX580.
With claymore (11.9), 186Mh/s @ 830W
With Phoenix (3.5d), 187Mh/s @ 1150W!
I tried to use the same values for clocks and voltages (which are set in config.txt) or to let it blank without any effect.
What am I doing wrong?

Install latest version of AMD drivers then switch to compute mode.

So you assume that he got 187 Mh/s from 6xRX580 with compute mode off? 
I think noobs account should not be allowed to give advice, only ask...

to answer OP: looks like clock and/or voltages are not properly applied, because power draw for phoenix is about the same as for claymore. And with that I refer you to  posts on this and previous page: "ALWAYS use 3rd party soft to control clocks, power and fans: OverdriveNtool for red and nvidiaInspector for green; don't use AB - it's shit. Don't use miners for controls."

[j2james]

Hi PM developers.

Today I have updated PM from 3.0c to 3.5d. And I cannot connect to CDM remote monitoring. I did not change -cdmport and -cdmpass.

I use command line parameters (not config file): -cdm 2 -cdmport 3333 -cdmpass xxxxx

Any ideas?

I have checked with config file. The same problem.

After back to 3.0c all works (with command line parameters and with config file)

   The only change with respect to remote monitoring in 3.5d is that now it starts 10-20 seconds after the miner starts but otherwise it should be working fine - we are using it 24/7 on our test rigs without any problems. From our own experience, this problem is perhaps related to Windows Firewall, which blocks the CDM port for the executable of the new version (it should show you a pop-up window, asking if you want to open the port the first time you start the a new version of PhoenixMiner). Another possible reason (again related to Windows Firewall) is that it may "decide" that the local network is "public" instead of "private" and it has much more strict rules for public networks, so it again blocks the CDM port of PhoenixMiner. A quick test to see if the Firewall is the culprit is to turn it off for a short time and see if the remote management works.

Hi. You are right. Thank you. Looks like I have ignored window of Windows Firewall this new run. I have fixed settings for PM in FW and all works.

[cbaa]

Hi there. I'm using phoenixminer since yesterday and till now it's nice. The only concern is the fact that fan speed cannot be controlled under linux. I'm waiting for your new version..

[Silverhydra57]

Hi,
Sorry if the problem has already been solved, I did not read the entire thread.

I am willing to give this miner a try but it draws far more power than claymore for a similar or little higher hashrate.

I use it with win10 with 6 RX580.
With claymore (11.9), 186Mh/s @ 830W
With Phoenix (3.5d), 187Mh/s @ 1150W!
I tried to use the same values for clocks and voltages (which are set in config.txt) or to let it blank without any effect.
What am I doing wrong?

Install latest version of AMD drivers then switch to compute mode.

So you assume that he got 187 Mh/s from 6xRX580 with compute mode off?  
I think noobs account should not be allowed to give advice, only ask...

to answer OP: looks like clock and/or voltages are not properly applied, because power draw for phoenix is about the same as for claymore. And with that I refer you to  posts on this and previous page: "ALWAYS use 3rd party soft to control clocks, power and fans: OverdriveNtool for red and nvidiaInspector for green; don't use AB - it's shit. Don't use miners for controls."

He used Blockchain driver.

That's right I use Blockchain drivers, I'll try with latest AMD drivers then and if that fails, I'll use OverdriNtool to control parameters (even if I prefer to apply it through miner).
Thanks a lot, you're both very helpful.

[Iamtutut]

So you assume that he got 187 Mh/s from 6xRX580 with compute mode off? 
I think noobs account should not be allowed to give advice, only ask...

to answer OP: looks like clock and/or voltages are not properly applied, because power draw for phoenix is about the same as for claymore. And with that I refer you to  posts on this and previous page: "ALWAYS use 3rd party soft to control clocks, power and fans: OverdriveNtool for red and nvidiaInspector for green; don't use AB - it's shit. Don't use miners for controls."

Afterburner works just fine (570s inside) and I have a wattmeter to check power usage at the wall.

[sxemini]

Hi,
Sorry if the problem has already been solved, I did not read the entire thread.

I am willing to give this miner a try but it draws far more power than claymore for a similar or little higher hashrate.

I use it with win10 with 6 RX580.
With claymore (11.9), 186Mh/s @ 830W
With Phoenix (3.5d), 187Mh/s @ 1150W!
I tried to use the same values for clocks and voltages (which are set in config.txt) or to let it blank without any effect.
What am I doing wrong?

Install latest version of AMD drivers then switch to compute mode.

So you assume that he got 187 Mh/s from 6xRX580 with compute mode off? 
I think noobs account should not be allowed to give advice, only ask...

to answer OP: looks like clock and/or voltages are not properly applied, because power draw for phoenix is about the same as for claymore. And with that I refer you to  posts on this and previous page: "ALWAYS use 3rd party soft to control clocks, power and fans: OverdriveNtool for red and nvidiaInspector for green; don't use AB - it's shit. Don't use miners for controls."

Why is afterburner shit? give reasons for your statement.
With afterburner you need only 1 programm to controll nvidia and amd. It is easy to use and to understand.

But agree with you: Don´t use miners for controls

Flash the best settings is the only right way.

[crypper]

Why is afterburner shit? give reasons for your statement.

Is personal experience and experience of many others shared on different forums good enough reason?

Flash the best settings is the only right way.

It's actually the worst thing you can do with your cards. And I guess you will ask for a reason: different algos require different clocks and even for the same algo different miners handle clocks differently.