The system I use is only for mining, so nothing other then software used for installing of preparing the GPU's is on that machine.. I always use DDU to do a full un install of the drivers and delete all (AMD) folders when there still are any left.. ( I have not cleaned the registry, as I thought DDU would do this for me).
I installed the 21.8.1 which are latest compatible drivers with 5.7b (as I read on page 1). I did the full monty install (as I need the AMD software to do the settings for the 6800 I have in the rig.).

It now runs for a while quite smooth again.. so as long as it runs I leave it.. ;-)
But thank you all for your comments... I understand now it is not normal to take that long for PM to start up...

Well PM 5.7b is finally running.. with the AMD 21.1.8.1
I know I am an impatient person..
With 5.5c it was about 10-20 seconds I had to wait before I could see the openCL drivers were loaded and the rest of the config file was executed.

but waiting for 5-6 minutes to finally see that the miner is actually start to load the openCL drivers and read the rest of the config file...
Is the long waiting period all have to do with the extra setting features in PM?

I understand it has something to do with the pentium (3,3Ghz) CPU I use, but from 10-20 sec to 5-6 minutes waiting is a bit much in my opinion.. Anything I can do to shorten this? without upgrading my CPU..
I also use 16Gb of RAM..

any help much appreciated.

Hi Deedeeranged, tnx for your reply. The settings as shown do work in PM 5.5c.
For a reason it seems that the -mvddc and -vmr does not work for the 5700 and 6800 cards.
The 6800 card I tune via the AMD software.

Do you think the missing data is the reason why PM 5.6b or 5.7b does not start??

After a long period of working with phoenixminer 5.5c and AMD drivers 20.1.1.1. and only some interventions the last month I had to restart the rig (AMD cards:4x 580, 1x 5700 3x 5700xt and a 6800) more often (driver errors).
So I thought it would be a good time to uninstall all and install new drivers (21.8.1) and use the newest phoenixminer 5.7b.

After the driver install I ran the phoenixminer. it starts to read the config file ( all up to showing the phoenixminer version (5.7b)

but it seems it doesn not load the drivers. (with 5.5c I saw for all 9 cards the driver version appearing)
I use the same items in the config file as for the 5.5c version.
What is /could be the problem why phoenixminer does not load the driver for the cards and start to mine?

thanks for the help.

I was searching for something like that, but could not find it. (so thank you for the link) I knew it was something aroud 150W... the thing is the 22Amps for the molex is, as I read an explanaition 11 amps per power cable so 11Amps for the 12V and also 11Amps for the 3.3V.. I also read that you can not add them.. so that makes up for the Watt and Amps...


Any thoughts about this? what could cause this?  Thank you.

Okay, I think our problem at first was that I connected to many risers to 1 SATA cable, but I also read online that SATA was kind of restricted to something like 9A and Molex near 22A. Therefore I believed it would be beter to have the risers connected to a molex cable and also connect not more than 2 risers per cable. Since I have two bad connectors now on my PSU I can not connect every riser to its own molex cable, I had to fit 2 on one cable. So far this seems to hold. (not overheat)
Soon I will have all the risers connected via PCI-e....

I came along another "challenge"..... After I saw a high number of rejected shares for only 1 card I altered (lowered) the VMR and wanted to restart the Phoenixminer, but when I did this it "noticed" a third (ghost) card (RX5700XT while we only have 2 actual RX5700XT in the rig) on position 2 in the list of cards and the number of usuable mining cards was also added with 1. ..... It does not show in the windows devicemanager list..

Any ideas what could have caused this?

 

After a while I have found the problem... SATA power for the risers... is not a good combo.. So after changing those the rig was stable..
Now with replacing some of the cards for newer ones we have new challenges... ;-)  Thanks for the support.. really appreciated.

Let me start with thanking you for your patience with me (and my friend)
After installing all available drivers of AMD (even the 21.3 which is not yet supported I found out ;-) ) it still keeps crashing

There are a few things I noticed...
With a few drivers the RX5700 card kept (re)setting its memclock to the number I set in the config file.
With last install of 20.11 drivers. I did not get a timeout on the driver (thread stuck....) but I saw an gigantic hashrate every one would like to get (from the logfile:  3: 2947100.674 MH/s) after a few of these line the log states:  "
"2021.03.17:11:59:05.099: GPU3 GPU3: Starting up... (0)
2021.03.17:11:59:05.621: hwmc GPU3: set GPU clocks to 1300 MHz (Vddc 850 mV)
2021.03.17:11:59:05.624: hwmc GPU3: set VMEM clocks to 1800 MHz"    <-- no vmddc mentioned

I no run the 5700 alone with following config:
" .........
-pool ssl://eu1.ethermine.org:5555
-pool2 ssl://us1.ethermine.org:5555
-wal --------------------------------------------
-prate 0.22
-cdm 2
#-cdmpass Bulldozer

-gpus 3
-tt 68
-fanmin 20
-tstop 77
-tstart 60
-amd
-acm
-cclock *:1150,rx*5700:1250 
-cvddc 850
-mclock rx*580:2175,rx*5700:1800
#-mclock 2175,2175,1800,2175,2175,2175,2175
-mvddc rx*580:855,rx*5700:950
#-mvddc 860,850,900,850,850,860,850
-gt 0
-vmr rx*580:29
-logdir --------------------------------
#-minrigspeed 200
-wdog 1
-rmode 2
......"
When starting Phoenixminer this is what the log recorded:
".......
2021.03.17:13:18:28.445: main Starting GPU mining
2021.03.17:13:18:28.445: main Matched GPU1 to ADL adapter index 48 (method 1)
2021.03.17:13:18:28.514: main GPU1: AMD driver 20.11.2
2021.03.17:13:18:28.514: main GPU1: Created ADL monitor for adapter 48; overdrive version: 8 (
2021.03.17:13:18:28.514: main GPU1: using AMD driver ver 20.11.2
2021.03.17:13:18:28.576: hwmc GPU1: set auto fan: 68C target temp (min fan 20, max fan 100)
2021.03.17:13:18:28.583: hwmc GPU1: set GPU clocks to 1250 MHz (Vddc 850 mV) (NO MEN clock or voltage)??
2021.03.17:13:18:28.646: wdog Starting watchdog thread
......"
In "stand alone" mode it is already running longer without a problem than before.
We have 2x750W of power.. 4 rx580 cards on PSU1 and 2x580, 1 5700 on PSU2. I think that should be sufficient.. we do about 1030W on the wallsocket.
What might cause the incredible super fantastic hashrates?

Well we tried a different (lower version) driver but then we had to change the kernel and still it didnt want to see the 5700 and the additions in the setting didnot seem to work, so this morning I changed to Win10 and we did see all the different cards.. and even the card settings in the confiig file (the RX*580:1160) did work. But we had the 21.1.1 drivers installed and it was not stable. So I uninstalled these and installed the 20.11.
and forced the compute mode via the AMD driver software, but I do not get it stable after a few minutes I get black screens and AMD report of a driver issue and then it all stops???

Tnx. I will check the data and the logfiles.
When I run the miner in the console as far as I know it always starts with the card names.. (which GPU are found for mining) In this list it shows 6 times the same (type) cards RX 580

I even had fitted the RX 5700 as a single available GPU, but then I get the miner error "NO GPU for mining"

Is there any known hardware limit for getting the RX 5700 XT's to work in Phoenixminer, e.g. minimum of RAM used, or anything else, which we might have overlooked?
The rig is fitted with only 4 Gb RAM.  (We use the Asrock H110 BTC Pro+ mobo)

Situation:

We are no experts miners..
First we used 6 Radeon RX 580  GPU's (mixed brands) with Ubuntu as OS and Phoenixminer 5.5c as miner for a while.. Then we added a 7th 580 which came available and with no hussle we got it running. I have been busy to get the best settings for each card and got decent hashrates..

Now we wanted to up the hashrates by replacing the 580's GPU's with RX 5700 XT's one by one.
But for some reason we do not get the first one to be recognized in Phoenixminer.
It shows in the PCI list of Linux..

To make sure the 580's would run kind of optimized and since I do not know yet at what GPU number the 5700 would be get by phoenixminer, I wrote following lines in the config file:

-cclock RX*580:1150
-cvddc RX*580:850

-mclock RX*850:2175
-mvddc RX*580:860

As I understood from the help for the config file this should set the core/mem clock and voltage for all GPU's that are named RX AND 580 with the specified numbers.. all other cards (the RX 5700 XT) should use their default.
But when the miner is started the settings are not used and the RX 5700 is not displayed and we start mining with only 6 iso 7 cards.

I tried to look for any sollutions already but came without any..

Any help, ideas are highly appreciated.