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

[PhoenixMiner]

please advice me. Now almost all vendor in my country cut off warranty if I flash bios so I have been use claymore with strap for a long time. after he look like stop update I start to try mining with pheonix miner but got a probelm.

I try -mt option. My asus 570 8gb (not flash bios) with mt 0 got 25MH, 4.0MH with -mt 1 and crash with -mt 2.

What am I done wrong? please advice me. I use latest amd driver.

    This is strange, as the -mt option works fine with most cards. We are going to implement more flexible straps for AMD cards soon (ideally - in the 5.2 version of PhoenixMiner) but in the meantime, you can try different drivers because the -mt option actually uses the straps from the driver.

Just reporting an issue I ran into on v5.1b with mixed-gen ETH mining on one XFX RX 5700 8GB + one XFX RX 480 8GB + one MSI RX 480 8GB on Ubuntu 20.04 (via both minerstatOS v1.5 and v1.6).  When I upgraded from PhoenixMiner v5.0e to v5.1b, I saw my RX 5700 hashrate drop to about 58% of normal, and the two RX 480s dropped to about 67% of normal hashrate.  Whether related to the miner or the OS, I cannot say, but I also saw the GPU memory clocks intermittently bouncing back to stock settings for 20 seconds at a time instead of locking in at my normal overclocks, and I would hear the fans adjust their throttles each time the clocks adjusted.

Rolling back to v5.0e followed by a reboot seems to have resolved these issues.  I have happily used a few versions of the PhoenixMiner for about four months on this mixed-gen rig without issue, so this was a bit of a surprise.  Unfortunately, my normal log file didn't have any suspicious messages or errors to report.

    This could be related to the OS and driver versions. We haven't tested on Ubuntu 20.04 yet (our Linux test rigs use Ubuntu 16.x and Ubuntu 18.x) and it is possible that the new AMD pro drivers for 20.04 are different that these for 18.04 (this is certainly true for 20.20-1098277, which is released only for 20.04 and is not supported by PhoenixMiner yet). We are going to build some rigs with 20.04 soon but this will only be reflected in the next release of PhoenixMiner.

Can you add memory strap support to the P106 series GPUs as well (P106-100, P106-90). They're basically just GTX 1060 but the mining version.

Here's the log output:

    First, you must run the miner as administrator in order to use the straps. It this doesn't help, we are also adding straps support for more Pascal-based cards for the next release (5.1c), which should be ready in a few days. As the number of different Pascal-based GPUs is too big to be able to obtain and test on all of the, we also added a new parameter in 5.1c that will allow you to "trick" PhoenixMiner that the card is supported even when it isn't, which should be useful for the more "exotic" GPU variants like Quadro, mining cards, etc.

Could someone explain what does this new strap command do? Does it worth to start experiment with it? Should we see better hashrate with it?

   It tightens the memory timings of GDDR5 or GDDR5X chips of the Pascal-based Nvidia cards (GTX10x0), which may lead to better hashrate. If your card is supported it is definitely worth it to try this option. For some cards (these with GDDR5X memory) the results are quite good but even for GDDR5 cards there will be some improvement. Note that for some cards like 1070 (but not 1070Ti), which suffer from lower hashrate with each new epoch, you may need to allow more power by using higher power limit or higher core clock in order to see the improvement with the current high DAG epoch numbers on ETH and ETC.

Hello,

I have two 5700xt mining together.
I would like to put differents fan settings like that, but it's not working, only one card is working the first one (-gpus 1) :

Code:

PhoenixMiner.exe -pool pool -wal wallet -worker worker 
-gpus 1 -gt 55 -tt 67 -cvddc 750 -cclock 1300 -fanmin 75 -fanmax 100 -mclock 1850 -amd -mode 1 -clKernel 1 -clNew 1 
-gpus 2 -gt 55 -tt 67 -cvddc 750 -cclock 1300 -fanmin 50 -fanmax 75 -mclock 1850 -amd -mode 1 -clKernel 1 -clNew 1

Is there a way to make different settings for each cards?

Thanks for any advice. 

   The way to specify different settings for each GPU is as follows:

Code:

PhoenixMiner.exe -pool pool -wal wallet -worker worker  -gt 55 -tt 67 -cvddc 750 -cclock 1300 -fanmin 50,75 -fanmax 75,100 -mclock 1850 -amd -mode 1 -clKernel 1 -clNew 1

   So, when the parameter is the same for all cards, just specify it once like thsi: -tt 67. When you want different values for each GPU, use it like this: -fanmin 50,75 - it this case the fanmin for the first card will be 50, for the second 75. Note that there must be no spaces between the comma and the numbers, and if you have more than two GPUs, you must specify a number for each of them.

do new features work under linux?

    Yes, both -ttli and the Nvidia memory timing adjustments work under Linux too. You must run the program as root in order to use the straps.

@Phoenixminer - if you can make a section of config file for delayed execution that would be good enough.
good job on timings integration!
adding kawpow could be even better, as this miner is mostly very stable:) 

     Not a bad idea but it may lead to instability. We will think about this for the next release. We are also exploring the possibility of adding KawPow support for the next release if time permits.

[BeefyBitcoinBoy]

Just reporting an issue I ran into on v5.1b with mixed-gen ETH mining on one XFX RX 5700 8GB + one XFX RX 480 8GB + one MSI RX 480 8GB on Ubuntu 20.04 (via both minerstatOS v1.5 and v1.6).  When I upgraded from PhoenixMiner v5.0e to v5.1b, I saw my RX 5700 hashrate drop to about 58% of normal, and the two RX 480s dropped to about 67% of normal hashrate.  Whether related to the miner or the OS, I cannot say, but I also saw the GPU memory clocks intermittently bouncing back to stock settings for 20 seconds at a time instead of locking in at my normal overclocks, and I would hear the fans adjust their throttles each time the clocks adjusted.

Rolling back to v5.0e followed by a reboot seems to have resolved these issues.  I have happily used a few versions of the PhoenixMiner for about four months on this mixed-gen rig without issue, so this was a bit of a surprise.  Unfortunately, my normal log file didn't have any suspicious messages or errors to report.

   This could be related to the OS and driver versions. We haven't tested on Ubuntu 20.04 yet (our Linux test rigs use Ubuntu 16.x and Ubuntu 18.x) and it is possible that the new AMD pro drivers for 20.04 are different that these for 18.04 (this is certainly true for 20.20-1098277, which is released only for 20.04 and is not supported by PhoenixMiner yet). We are going to build some rigs with 20.04 soon but this will only be reflected in the next release of PhoenixMiner.

Wow, thanks very much for the reply, @PhoenixMiner.  For clarification, this also occurred under Ubuntu 18.04.3 LTS (bionic).  The 18.04.3 build is on amdgpu-pro 19.50-967956 drivers.  For both builds (Ubuntu 18.04.3 and Ubuntu 20.04), rolling back to PhoenixMiner v5.0e resolved the issue.  Thanks very much for the heads-up that Ubuntu 20.04 is not supported, yet!  I'll keep investigating and let you know if I find anything else that might be useful.

[guyom78]

Hello,

I have two 5700xt mining together.
I would like to put differents fan settings like that, but it's not working, only one card is working the first one (-gpus 1) :

Code:

PhoenixMiner.exe -pool pool -wal wallet -worker worker 
-gpus 1 -gt 55 -tt 67 -cvddc 750 -cclock 1300 -fanmin 75 -fanmax 100 -mclock 1850 -amd -mode 1 -clKernel 1 -clNew 1 
-gpus 2 -gt 55 -tt 67 -cvddc 750 -cclock 1300 -fanmin 50 -fanmax 75 -mclock 1850 -amd -mode 1 -clKernel 1 -clNew 1

Is there a way to make different settings for each cards?

Thanks for any advice. 

   The way to specify different settings for each GPU is as follows:

Code:

PhoenixMiner.exe -pool pool -wal wallet -worker worker  -gt 55 -tt 67 -cvddc 750 -cclock 1300 -fanmin 50,75 -fanmax 75,100 -mclock 1850 -amd -mode 1 -clKernel 1 -clNew 1

   So, when the parameter is the same for all cards, just specify it once like thsi: -tt 67. When you want different values for each GPU, use it like this: -fanmin 50,75 - it this case the fanmin for the first card will be 50, for the second 75. Note that there must be no spaces between the comma and the numbers, and if you have more than two GPUs, you must specify a number for each of them.

Thank you so much <3 You made my life easier.

[Guigs321]

Hello everyone !

I have a rig with this error : "FATAL ERROR : Debugger Detected" just after the listing of the graphics cards.

Any idea why ?

RIG : 6x RX470 4GB, Windows 10 up to date, Adrenalin 20.2.2, PM 5.1b

I have the same error with previous miner version, same with ETH, ETC or Nicehash...

Code:

2020.07.20:11:38:39.592: main Phoenix Miner 5.1b Windows/msvc - Release build
2020.07.20:11:38:39.592: main Cmd line: -pool stratum+tcp://daggerhashimoto.eu.nicehash.com:3353 -wal xxx -pass x -proto 4 -stales 0 -GT 37 -mport -3333 -timeout 1440 -rmode 1 -mi 14 -hstats 2 -dagrestart 2 -rvram -1 -cclock 1150 -mclock 2060 
2020.07.20:11:38:39.593: main No CUDA driver found
2020.07.20:11:38:44.652: main Available GPUs for mining:
2020.07.20:11:38:44.652: main GPU1: Radeon (TM) RX 470 Graphics (pcie 1), OpenCL 2.0, 4 GB VRAM, 32 CUs
2020.07.20:11:38:44.652: main GPU2: Radeon (TM) RX 470 Graphics (pcie 2), OpenCL 2.0, 4 GB VRAM, 32 CUs
2020.07.20:11:38:44.652: main GPU3: Radeon (TM) RX 470 Graphics (pcie 3), OpenCL 2.0, 4 GB VRAM, 32 CUs
2020.07.20:11:38:44.652: main GPU4: Radeon (TM) RX 470 Graphics (pcie 7), OpenCL 2.0, 4 GB VRAM, 32 CUs
2020.07.20:11:38:44.652: main GPU5: Radeon (TM) RX 470 Graphics (pcie 10), OpenCL 2.0, 4 GB VRAM, 32 CUs
2020.07.20:11:38:44.652: main GPU6: Radeon (TM) RX 470 Graphics (pcie 11), OpenCL 2.0, 4 GB VRAM, 32 CUs
2020.07.20:11:38:44.655: main ADL library initialized

[maxime506]

please advice me. Now almost all vendor in my country cut off warranty if I flash bios so I have been use claymore with strap for a long time. after he look like stop update I start to try mining with pheonix miner but got a probelm.

I try -mt option. My asus 570 8gb (not flash bios) with mt 0 got 25MH, 4.0MH with -mt 1 and crash with -mt 2.

What am I done wrong? please advice me. I use latest amd driver.

    This is strange, as the -mt option works fine with most cards. We are going to implement more flexible straps for AMD cards soon (ideally - in the 5.2 version of PhoenixMiner) but in the meantime, you can try different drivers because the -mt option actually uses the straps from the driver.

Just reporting an issue I ran into on v5.1b with mixed-gen ETH mining on one XFX RX 5700 8GB + one XFX RX 480 8GB + one MSI RX 480 8GB on Ubuntu 20.04 (via both minerstatOS v1.5 and v1.6).  When I upgraded from PhoenixMiner v5.0e to v5.1b, I saw my RX 5700 hashrate drop to about 58% of normal, and the two RX 480s dropped to about 67% of normal hashrate.  Whether related to the miner or the OS, I cannot say, but I also saw the GPU memory clocks intermittently bouncing back to stock settings for 20 seconds at a time instead of locking in at my normal overclocks, and I would hear the fans adjust their throttles each time the clocks adjusted.

Rolling back to v5.0e followed by a reboot seems to have resolved these issues.  I have happily used a few versions of the PhoenixMiner for about four months on this mixed-gen rig without issue, so this was a bit of a surprise.  Unfortunately, my normal log file didn't have any suspicious messages or errors to report.

    This could be related to the OS and driver versions. We haven't tested on Ubuntu 20.04 yet (our Linux test rigs use Ubuntu 16.x and Ubuntu 18.x) and it is possible that the new AMD pro drivers for 20.04 are different that these for 18.04 (this is certainly true for 20.20-1098277, which is released only for 20.04 and is not supported by PhoenixMiner yet). We are going to build some rigs with 20.04 soon but this will only be reflected in the next release of PhoenixMiner.

Can you add memory strap support to the P106 series GPUs as well (P106-100, P106-90). They're basically just GTX 1060 but the mining version.

Here's the log output:

    First, you must run the miner as administrator in order to use the straps. It this doesn't help, we are also adding straps support for more Pascal-based cards for the next release (5.1c), which should be ready in a few days. As the number of different Pascal-based GPUs is too big to be able to obtain and test on all of the, we also added a new parameter in 5.1c that will allow you to "trick" PhoenixMiner that the card is supported even when it isn't, which should be useful for the more "exotic" GPU variants like Quadro, mining cards, etc.

Could someone explain what does this new strap command do? Does it worth to start experiment with it? Should we see better hashrate with it?

   It tightens the memory timings of GDDR5 or GDDR5X chips of the Pascal-based Nvidia cards (GTX10x0), which may lead to better hashrate. If your card is supported it is definitely worth it to try this option. For some cards (these with GDDR5X memory) the results are quite good but even for GDDR5 cards there will be some improvement. Note that for some cards like 1070 (but not 1070Ti), which suffer from lower hashrate with each new epoch, you may need to allow more power by using higher power limit or higher core clock in order to see the improvement with the current high DAG epoch numbers on ETH and ETC.

Hello,

I have two 5700xt mining together.
I would like to put differents fan settings like that, but it's not working, only one card is working the first one (-gpus 1) :

Code:

PhoenixMiner.exe -pool pool -wal wallet -worker worker 
-gpus 1 -gt 55 -tt 67 -cvddc 750 -cclock 1300 -fanmin 75 -fanmax 100 -mclock 1850 -amd -mode 1 -clKernel 1 -clNew 1 
-gpus 2 -gt 55 -tt 67 -cvddc 750 -cclock 1300 -fanmin 50 -fanmax 75 -mclock 1850 -amd -mode 1 -clKernel 1 -clNew 1

Is there a way to make different settings for each cards?

Thanks for any advice. 

   The way to specify different settings for each GPU is as follows:

Code:

PhoenixMiner.exe -pool pool -wal wallet -worker worker  -gt 55 -tt 67 -cvddc 750 -cclock 1300 -fanmin 50,75 -fanmax 75,100 -mclock 1850 -amd -mode 1 -clKernel 1 -clNew 1

   So, when the parameter is the same for all cards, just specify it once like thsi: -tt 67. When you want different values for each GPU, use it like this: -fanmin 50,75 - it this case the fanmin for the first card will be 50, for the second 75. Note that there must be no spaces between the comma and the numbers, and if you have more than two GPUs, you must specify a number for each of them.

do new features work under linux?

    Yes, both -ttli and the Nvidia memory timing adjustments work under Linux too. You must run the program as root in order to use the straps.

@Phoenixminer - if you can make a section of config file for delayed execution that would be good enough.
good job on timings integration!
adding kawpow could be even better, as this miner is mostly very stable:) 

     Not a bad idea but it may lead to instability. We will think about this for the next release. We are also exploring the possibility of adding KawPow support for the next release if time permits.

Thank you Phoenix Miner! Unfortunately running it at admin level doesn't work either. I'll wait for next release

[Ursul0]

@Phoenixminer - if you can make a section of config file for delayed execution that would be good enough.
good job on timings integration!
adding kawpow could be even better, as this miner is mostly very stable:) 

     Not a bad idea but it may lead to instability. We will think about this for the next release. We are also exploring the possibility of adding KawPow support for the next release if time permits.

it seems stable enough (I was able to init only with straps 0, and then switched to 5):




Also I tested on 1060 - the results are mixed. its 10*1060*6G, with this config:
-mclock +800,+800,+800,+775,+800,+775,+775,+800,+775,+800
-powlim -36,-36,-36,-36,-36,-36,-36,-36,-36,-36
-cclock +100,+100,+100,+100,+100,+100,+100,+100

straps 1 I can apply before the dag generation, and the speed goes from 234mh to about 240+ then 238 after few hours, then back to the initial 233-234. with other straps applied after the dag generation I can get better results, but the improvement seem to degrade with time... what gives?

[greasepit]

I just woud like to register here my ethash rate on my 24x Asus (Arez) RX 580 8GB

30.1 Mh/s stable
https://i.ibb.co/0cVRbBT/hash-rate.png

Rigs
https://i.ibb.co/q7T8FNM/RX-580-8-GB.png
https://i.ibb.co/xJ2p6JQ/RX-580-8-GB-1.png


-tmax 69
-fanmin 70
-tstop 75
-tstart 40
-minRigSpeed 155
-hstats 2
-cdmpass ********
-mt 1
-cclock 1100
-cvddc 800
-mclock 2050
-mvddc 825
-rmode 2

I did some modifications on the timmings like:
--R2W 31 --RC 75 --RRD 5 --RCDRA 26 --RCDR 26 --RCDWA 14 --RCDW 14 --RFC 184 --RP_WRA 55 --RASMACTWR 33 --RASMACTRD 28 --ACTWR 9 --ACTRD 10 --RAS2RAS 101 --RP 24 --WRPLUSRP 30 --REF 45

PhoenixMiner 5.0e
AMD driver 20.4.2
Windows 10 19041.264

I still need to finish installing the other cards
Using 4x 220v 23w FAN at the back of each the rig (too loud)

hello
how set the timings ? read the readme and nothing found i try google some forums and nothing my best speed my rx 580 25 mhs please help
thank you

I tweaked the timings using a third party software called AMD Memory Tweak you can find more information on this topic https://bitcointalk.org/index.php?topic=5123724.0

[neogeo]

Hello.
Need some HELP about -ppf command.

How to set those command per GPUs because i have 5 GPUs and want to dedicate different -ppf value per each GPU?

TNX

(Cards are rx 570s and rx 580s)

[UnclWish]

Hello.
Need some HELP about -ppf command.

How to set those command per GPUs because i have 5 GPUs and want to dedicate different -ppf value per each GPU?

TNX

(Cards are rx 570s and rx 580s)

Just read the theme.

Also I tested on 1060 - the results are mixed. its 10*1060*6G, with this config:
-mclock +800,+800,+800,+775,+800,+775,+775,+800,+775,+800
-powlim -36,-36,-36,-36,-36,-36,-36,-36,-36,-36
-cclock +100,+100,+100,+100,+100,+100,+100,+100

The same way.

[neogeo]

Can you help me specifically or just share demagogic advices?

[UnclWish]

Can you help me specifically or just share demagogic advices?

It's not demagogic. Any parameter can be set like
-mclock 1250
for all devices or
-mclock 1250,0,1100 for each device individually. 1250 for 1st GPU, didn't change for 2nd GPU, 1100 for 3rd GPU etc.

Ask what you want to set for your devices and i can help you if you can't do that.

[neogeo]

I know whole thing, and a single command -ppf 65 working on whole rig from 5 cards but if i set it up as -ppf 65,66,67,65,65 then miner/rig don't working (don't working when have commas).
That's my problem

P.S. Sorry for my nervousness

[igotek]

v5.1b 

I see new version on HiveOS. What is the new features ?

[mogielev]

Hi there!
I have problem today: after start Phoenix Miner 5.1b (and 5.0e before) miner was crash. This happened today about 05:00 (GMT +2) simultaneously on 2 rigs: they was work and suddenly crash:
https://piccy.info/view3/13906056/1e802ada36006862d3b9e01132e670d0/

I try restart miner, but this also crash:
https://piccy.info/view3/13906061/b236c788f5aef49139a7709d0616fc38/orig/
Crash happens after Light cache generated in as you see on the picture.

This same on the Phoenix Miner 5.1b and Phoenix Miner 5.0e. I dont do anything: rigs work and suddenly crash and now i cant run they.

Have you something idea about this?

[UnclWish]

I know whole thing, and a single command -ppf 65 working on whole rig from 5 cards but if i set it up as -ppf 65,66,67,65,65 then miner/rig don't working (don't working when have commas).
That's my problem

P.S. Sorry for my nervousness

What error miner gives? Give screenshot.
Try to check by this way:
-ppf 65,65,65,65,65

[extraspicywasabi]

Hi Phoenix Miner DEVs

Epoch 350 is now the reported DAG on my miner.

Running 5.0e on Windows 10 with the latest updates, etc.,

Have been having issues, only noticed today (so I'm not sure how long it's been happening, or if you already know about the issue?).  The hash rate has dropped noticably, and there is an error running one of the commands.

----------------------------------

hardware:

x2 AMD Radeon RX470 8GB Mining Edition with the latest AMD Drivers 20.7.2
x1 AMD Radeon RX57000XT 8GB  with the latest AMD Drivers 20.7.2

----------------------------------

prior to Epoch 350:

RX470, 30.3 MH/s, consistent rate
RX5700XT, 54 MH/s, consistent rate

Total 114.6 MH/s

----------------------------------

since Epoch 350:

RX470, 29.3 MH/s, consistent rate
RX5700XT, 48 MH/s - 49.5 MH/s, variable rate

Total 106.6 MH/s - 108.1 MH/s, variable rate

(54 MH/s a bit low for an RX5700XT, but the GFX cards are all squashed in a desktop PC case and this card in particularly is being throttled by thermals, since I'm aware of this though, that's fine from my point of view until I pickup another case or open air mining frame)

----------------------------------

The AMD Radeon RX470 8GB Mining Edition, I was running the cards with the below argument, which worked prior to Epoch 350:

-clKernel 3

GPU1: clSetKernelArg (-49)
Fatal error detected. Restarting

----------------------------------

Running with the optimised setting prevents the fatal error:

-clKernel 1

----------------------------------

Since I was using the -clKernel 3 argument prior to Epoch 350 without an issue, and the argument produced the best/optimal hashrate for my RX470 GPUs and now there is an error and slightly lower overall hashrate, this indicates there is a bug somewhere, and I wanted to report.  I tested the same config on Phoenix 5.0b which had the same result. Other people may not report, and simply use other miners, and you'll lose their DEV FEEs as a result.

[patthara74]

Since 21/07 morning got a issue with rx5700 look like it can't reduce core clock to setting, last night still fine.

Update: reflash again now I can reduce core speed to setting but speed reduce from 56.8 to 50.5 T_T

[asarhad]

i tried to start miner today, and after initial lines of iniialising cards, after line generated light it just stop without any error both in CMD or in log.
6x580 4Gb cards and win10 with BETA version of drivers 17.x

any idea why miner just ends without any error/note/


regards

[4play]

I have win 10 and 6x 1070 GPUs, and i using last PM version

Hasrate contanstly falling. Until today I using nvidia drivers 388.43 today I change (update) drivers to 442.59. Hasrate is still falling. My OC settings are
https://imgur.com/wj5jLir.png

Current PM settings:

PhoenixMiner.exe -pool eth-eu1.nanopool.org:9999 -pool2 eth-eu2.nanopool.org:9999 -wal ETHaddress.rig01/mymail@mymail.me -epsw x -ftime 10 -rmode 2 -lidag 2 -wdog 1 -astats 1 -proto 2 -logsmaxsize 1

Does anyone have any idea why hasrate falling. Last 3 maybe 4 months I have avg. hasrate aournt 187 to 189 MH/s 6 months ago constantly 194.7 MH/s.

Tnx for all you answers and tips.

[Digital_Seytan]

Phoenixminer 5.1b is a problem at all, 3x rx 570 4gb keeps crashing, and hashrate normally 3016 now 2916 better even older versions phoenixminer, and why you can't mine nimiq with phoenixminer while using temredminer very stable eth and nimiq minen super..


https://github.com/todxx/teamredminer/releases


Happy mining :0