Quote from: majika on November 16, 2021, 06:40:21 PM

Sup, So PhoenixMiner ver's 5.6-5.8 are driving me crazy.. (Just about to test 5.9beta)

MY QESTION IS: Which is the best drivers to use for each of the Phoenix Miner Releases in order from Ver 5.5 onwards please..?

- A Follow up question would be, What is the best most reliable version of drivers to use on the 5.9beta release? Will I be safe with Adrenalin 21.10.2 OR should I go back to Adrenalin 20.8.1 ??

RIG SPECS

CPU: i7-7700K
RAM: 8GB
MOBO: ASUS B250 Mining Expert - Loaded Bios Optimised Defaults. Pcie link > Gen 1, etc..
PSU: x2 EVGA 1600w T2 Platinum/Titanium (I forget)
RAISERS: I just replaced All my GPU Raisers to the newer 009s Plus BLACK EDITION
OS: Win10 x64 Pro [Clean format]
GPUs RX580 8GB: x2 RX580 8GB Nitro+ (Hynix), x6 Nitro SE RX580 8GB (Samsung), x1 Power Devil RX580 8GB (Micron), x1 Asus ROG RX580 8GB (Hynix)

**Current** Drivers: Adrenalin 19.8.2 Beta

Tweeks: Applied Win10Tweeks.bat and done all the usual tweaks (Win update OFF etc..)

Set: Windows ENV_MEM Initial 60000mb > 88000mb

I am trying to pin down why my rig RXRig (x10 RX580's) keeps throwing errors and/or having the following issues:

I have managed to have a run of about 6hrs then another run for like 12hrs But crashing/BSOD not able to auto_restart miner from shell:startup > miner_start.bat

** Timeout Detection Recovery atikmpag.sys / atikmdag.sys VIDEO_TDR_FAILURE - TDS_BSOD errors.?

I had noticed a big jump in Temps, causing -tstop or -ttli to kick in.. Usually, my cards idle around hi 50's - to low - 60C deg C. (-tt 60)
** Higher than normal stale shares.
** -clKernel 1
** I use -altinit and -mi 12

** The issues began on v5.8c

I have two separate config's called from 2 different batch scripts depending on time of year/temperature outside

1) Standard config -mt 0
2) Winter (RXBoost config) config -mt 0 -vmr 25 -strap 1 -vmdag 1

All cards are now on -cclock 1150/850mv-900mv -mclock-2100-2150:900-925mv
I have down clocked my cards and undervolted

I was initially was running a stable -mclock 2170 on gpu's Samsung/Hynix/Micron cards (Mixed) ALL Were stable @ -mclock 2170 between 875-925mv voltages, with a range of -cclocks 1150,1170,1240,1150... etc... STABLE upto version PhoenixMiner v5.8

Now I have formatted my Rig to clean slate. I am running through all the various things I can think of to figure out why my rig keeps crashing.
Current testing on PNX Miner v5.6d - > 5.7 with Adrenalin 19.8.2 Beta drivers..

If still no luck I will bump up to last stable set of drivers 20.8.1 and repeat the testing steps. But as it stands. v5.8 really threw a wrench in the rig and cannot get it to run stable now at all..
Any suggestions would be appreciated...

Cheers Maj

I see that you use -mt 0, which if I understand correctly is for the original bios, not for preflash ones. These are my settings with fashionable bios and driver 21.3.1
-acm -amd -vmr 75 -gt 67,34,34,34,34,64 -mi 13 -eres 0 -clKernel 2 -gpus 1,2,4,5,6,7 -stales 0 -mclock 2250,2070,2085,2095,2070,2250 -cclock 1250 -cvddc 900 -mvddc 900

slava9722

Posts

Merits