SRBMiner Cryptonight AMD GPU Miner V1.9.3 - native algo switching

[doktor83]

Try with kernel 2 it will work

next version will set it automatically

That worked.  Thanks!

Did you use kernel 2 only for the r9 200 series or you had to use them on the R9 390 Series also?

edit: have you tried int: 30 , w: 8 , threads: 3 setup on your hawaii's ?

[1714063600]

1714063600

[1714063600]

1714063600

[1714063600]

1714063600

[zeef]

I have a request, I know there have been discussions recently, but I am one of those that use profit switchers, and I would like you to kindly make your devfee mine more often and for less seconds to keep at the same level. Hell, I could even be glad with 1% devfee. Right now I am using a cooldown of 2h on my profitswitcher so it doesn't mine devfee for every miner restart and reach 5%+ devfee, but I am missing opportunities to mine the most profitable coin (in 2h, difficulty adjusts). Please, please make it something like every 30 minutes at least (it can mine the first devfee in the first 15 minutes, like it is now). I couldn't care less if you make it not show in the screen...I know you're fair and trustworthy, but please take my feedback into consideration.

How often (approx.) does your profit switcher restarts miner ? 5, 10, 15, 60 min?
Have you guys heard about Monero Ocean pools?

You use your XMR address as login, but the pool switches to the most profitable algos, so you can mine the most profitable algo/coin, but you are payed in XMR.

I integrated everything and it is working, so if i decide so, SRB could support Monero Ocean algo switching in next version.
Two problems at the moment : miner has to restart when changing algo (ATM it can't change algo on the fly), the devfee mining.

It can switch algos every 5 minutes, but sometimes does not switch for hours, depending what is the most profitable to mine at that moment.
I just have to think of a way to get the devfee, but at the same time not force the user to pay a lot more fee (like when you restart the miner every 20 minutes, right ? )

Do it do it hehe 

[RadikalCID]

Try with kernel 2 it will work

next version will set it automatically

That worked.  Thanks!

Did you use kernel 2 only for the r9 200 series or you had to use them on the R9 390 Series also?

edit: have you tried int: 30 , w: 8 , threads: 3 setup on your hawaii's ?

In my case for the GPUs R9 390 8GB - R9 390X 8GB (Hawaii Cards) always the best option is Threads: 3

[wgestickley]

Please implement ocean to SRB Miner!

[doktor83]

Please implement ocean to SRB Miner!

Soon 

[wgestickley]

Please implement ocean to SRB Miner!

Soon 

Perfect 👌

Does that mean full Algo support with all by MoneroOcean supported Coins and Algos?

Or a “light” Version like the actual Cast-XMR or XMR-Stak versions, where it can only switch to the profitabel Coins in one specified Allgo?

[doktor83]

Please implement ocean to SRB Miner!

Soon 

Perfect 👌

Does that mean full Algo support with all by MoneroOcean supported Coins and Algos?

Or a “light” Version like the actual Cast-XMR or XMR-Stak versions, where it can only switch to the profitabel Coins in one specified Allgo?

Full algo switching support yes.

[wgestickley]

Please implement ocean to SRB Miner!

Soon 

Perfect 👌

Does that mean full Algo support with all by MoneroOcean supported Coins and Algos?

Or a “light” Version like the actual Cast-XMR or XMR-Stak versions, where it can only switch to the profitabel Coins in one specified Allgo?

Full algo switching support yes.

Nice 👍

And what means „soon“ in your case?
I am very excited about it 😀

[Privatoria]

You have very interesting project especially now, when mining profitability is getting down. I read your FAQ about rendering but I didn’t find all answers how to participate in your project, what is exact amount of tokens do I need to have to participate. I hope that after pre-sale you’ll have more detailed guide on your site.

[ViWe]

You have very interesting project especially now, when mining profitability is getting down. I read your FAQ about rendering but I didn’t find all answers how to participate in your project, what is exact amount of tokens do I need to have to participate. I hope that after pre-sale you’ll have more detailed guide on your site.

Wrong post dude.

[azzalea25]

SRBMiner Cryptonight AMD GPU Miner
http://www.srbminer.com
-----------------------------------------------------------------------------------------------
Windows only

The FASTEST Cryptonight miner for Vega cards you can get


Download V1.6.5:

Direct: http://www.srbminer.com/download.html
Mega : https://mega.nz/#F!qVIgxAwB!kKmgCDICmQwbdVvMb-tAag

MD5  : 375cbb1afe5a3ecf782c53d5452665d0 *SRBMiner-CN-V1-6-5.zip

Download only from these links, these are the only official links

Use with Adrenalin 18.6.1 or Adrenalin 18.3.4 drivers!
Don't use Adrenalin 18.7.1 drivers



!!!!!!!!!
If you are a beginner miner and need some help in setting up SRBMiner, check out THIS LINK
!!!!!!!!!



Supported algos:


- Cryptonight [normal]
- Cryptonight Lite [lite]
- Cryptonight V7 [normalv7]
- Cryptonight Lite V7 [litev7]
- Cryptonight Heavy [heavy]
- Cryptonight Haven [haven]
- Cryptonight Fast [fast]
- Cryptonight BitTubeV2 [bittubev2]
- Cryptonight StelliteV4 [stellitev4]
- Cryptonight ArtoCash [artocash]
- Cryptonight Alloy [alloy]
- Cryptonight B2N [b2n]
- Cryptonight MarketCash [marketcash]
- Cryptonight Italo [italo]
- Cryptonight Red [mox]



Supports Nicehash & SSL/TLS encrypted connections



DevFee:
- Low DevFee (0.85%) [First devfee is mined randomly in the first 15 minutes, the rest is every 2 hours]
- Non-agressive DevFee mining -> if miner can't connect to DevFee pool, no problem, switching back to user pool ASAP


Extra:
- Watchdog that monitors your GPU threads, if they stop hashing for a few minutes, miner restarts itself
- Hash monitor, if 5 minute average hash falls under the value you define, miner restarts itself
- Set system shutdown temperature, to protect your GPU's from overheating
- Restart (disable/enable) Vega gpu's before mining for maximum performance
- API for rig monitoring
- Set compute mode and disable crossfire on all cards
- Benchmark every algo locally without connecting to a pool



Here's a (not complete) list of supported coins, and which algo has to be used

normal
+Electroneum
+Monero Classic
+ByteCoin
+Karbo
+DigitalNote
+Leviar
+Ultranote
+BalkanCoin
+BitSum
+CrepCoin
+DinastyCoin
+LeviarCoin
+BitCoal
+Sumo
+IntuCoin
+Geldum
+Newton
+X12Coin

normalv7
+Monero
+MoneroV
+Intense
+Graft
+BBS
+Wownero
+Lines
+Fonero
+Myztic
+Catalyst
+Quantum Resistant Ledger
+ReCoal
+xDrop
+Leviar

lite

litev7
+Turtlecoin
+Iridium
+Aeon
+Arqma

heavy
+Ryo
+Loki
+Bloc.Money
+BixBite
+Saronite
+Ombre
+ItaloCoin
+Niobio
+ZBucks

bittubev2
+Ipbc/BitTube

artocash
+Artocash

alloy
+Alloy

marketcash
+MarketCash

b2n
+Bitcoin2Network

stellitev4
+Stellite

fast
+Masari
+Conceal
+PrivatePay
+Kepl

haven
+Haven

italo
+ItaloCoin

mox
+MoX Project


--------------------------------------------------------------------------
CONFIG.TXT
--------------------------------------------------------------------------


You can use these options :
Note: use all lowercase characters

"cryptonight_type" : "normal, normalv7, lite, litev7, heavy, bittubev2, artocash, alloy, marketcash, b2n, stellitev4, haven, fast, italo"
"intensity" : 0-300, if set to 0 miner will try to find best settings for every video card
"worksize" : 1-256, if set, every video card will use this worksize, if not set, using auto detected value
"double_threads" : true or false, set it to true for best performance
"giveup_limit" : number, how many times to try connecting to a pool before switching to next pool from pools.txt
"timeout" : number - seconds, when is a connection to a pool treated as time out
"retry_time" : number - seconds, how much to wait before trying to reconnect to a pool
"reboot_script" : filename to a batch file in miner directory, if set it turns off built in miner reset procedure on gpu failure, and instead runs this script
"restart_devices_on_startup" : true or false, if true it will use devcon to disable/enable all Vega cards before mining
"restart_devices_on_startup_script" : filename to a batch file in miner directory, this script will run after disable/enable and only if 'restart_devices_on_startup' is true. You can use it to set OverdriveNTool for example.
"main_pool_reconnect" : number - seconds (minimum is 3 minutes or 180 sec), how often to try to reconnect back to the main pool. Default is 10 minutes.
"min_rig_speed" : number - in H/S, it defines the minimum rig hashing speed we want to maintain. If 5 minute average hashing speed is less than this, miner restarts.
"min_rig_speed_duration" : number - in seconds (minimum is 30 sec), defines the time period for average hashing speed calculation, used with 'min_rig_speed' parameter

"api_enabled" : true or false, if true we can connect to it on 127.0.0.1:21555
"api_rig_name" : identifier name for your rig
"api_port" : number, this is where API will be available (Default is 21555 if not set)

"target_temperature" : number between 0-99, miner will try to maintain this temperature on all found video cards (ADL must be enabled, works only on cards supporting OverdriveN)
"shutdown_temperature" : number between 0-100, if this temperature is reached, miner will shutdown system (ADL must be enabled)


DON'T comment out or remove intensity and double_threads setting at the top of the config file, just leave it as-is.



=================================
Manual GPU setup (advanced)
=================================

To manually set up video cards, you must create a "gpu_conf" array in the config.txt file.

Example :

"gpu_conf" :
[
   { "id" : 0, "intensity" : 50, "worksize" : 4, "threads" : 1},
   { "id" : 1, "intensity" : 45, "worksize" : 4, "threads" : 2},
   { "id" : 2, "intensity" : 40, "worksize" : 8, "threads" : 3},
   { "id" : 3, "intensity" : 55, "worksize" : 8, "threads" : 4}
]


Additional parameters:

"kernel" : 0-2 , if not set, or set to 0, miner will select most suitable kernel (1-FOR GCN CARDS, 2-FOR PRE-GCN CARDS)
"target_temperature" : 0-99, if set miner will try to maintain this temperature for this particular video card. If option 'target_temperature' on top of config.txt is set, this option WILL BE IGNORED. (ADL must be enabled, works only on cards supporting OverdriveN)
"target_fan_speed" : 0-6000, if set miner will try to set the video card fan speed to this speed. Setting is in RPM (rounds per minute) (ADL must be enabled)
"adl_type" : 1 or 2 , 1 - USE OVERDRIVEN , 2 - USE OVERDRIVE 5. Default is 1 if not set. Option 2 (Overdrive 5) is suitable for older cards
"persistent_memory" : true or false, if set miner will try to allocate extra memory for the video card, if it is available. CAUTION, MINER CAN BECOME UNSTABLE AND CRASH if using this option




--------------------------------------------------------------------------
POOLS.TXT
--------------------------------------------------------------------------


Example:

{
"pools" :
[
   {"pool" : "pool_1_address", "wallet" : "pool_1_wallet", "password" : "x"},
   {"pool" : "pool_2_address", "wallet" : "pool_2_wallet", "password" : "x"},
   {"pool" : "pool_3_address", "wallet" : "pool_3_wallet", "password" : "x"}
]
}

Pool on first position is the MAIN pool, others are counted as FAILOVER pools.

Additional parameters:

"worker" : worker name, not every pool supports this
"nicehash" : true or false, set this to true if you are using Nicehash
"keepalive" : true or false, not every pool supports this
"pool_use_tls": true or false, if true miner will use SSL/TLS to connect to pool
"job_timeout" : number in seconds, if no job is received for this period, miner will reconnect to the pool (Default is 20 minutes)
"max_difficulty" : number, if pool difficulty is above this value miner will reconnect to the pool



--------------------------------------------------------------------------
START.BAT
--------------------------------------------------------------------------

There are some options that must be set in start.bat, and not within config.txt or pools.txt.
For example if you want to create configurations for multiple coins, you must define a 'start.bat' like file for every coin.

Parameters:

--config filename (use config file other than config.txt)
--pools filename (use pools file other than pools.txt)
--logfile filename (enable logging to file)
--listdevices (list available devices)
--listdevicesreordered (list available devices ordered by busid)
--gpureorder (order devices by busid)
--adldisable (disable ADL)
--disablegpuwatchdog (disable gpu crash detection watchdog)
--resetfans (reset fans back to default settings on miner exit)
--enableduplicategpuid (enables the usage of the same gpu id in gpu_conf multiple times)
--sendallstales (send every stale share - may increase number of rejected shares)
--setcomputemode (sets AMD gpu's to compute mode & disables crossfire - run as admin)
--runbenchmark (benchmark your current algo settings offline)
--benchmarkduration (how long to run the benchmark in seconds, max is 3 minutes, def. is 1 min.)


When setting any of the parameters below, don't use " or ' around the value!


To setup your video cards in cmd line :

These settings override the settings in config.txt
If you want to set everything in cmd line, you still need to have an empty config.txt file (which contains just : {}, or any other parameter like api etc etc )
First list devices (--listdevices or --listdevicesreordered if you are going to use --gpureorder), then you know the GPU id's and can set them up easy.

Parameters:

--ccryptonighttype value (algo to use)
--cgpuid value (gpu id, comma separated values, use --listdevices to see available)
--cgpuintensity value (gpu intensity, comma separated values)
--cgputhreads value (number of gpu threads, comma separated values)
--cgpuworksize value (gpu worksize, comma separated values)
--cgputargettemperature value (gpu temperature, comma separated values)
--cgputargetfanspeed value (gpu fan speed in RPM, comma separated values)
--cgpuadltype value (gpu adl to use (1 or 2), comma separated values)
--cgpukernel value (gpu kernel to use (1 or 2), comma separated values)



To setup your main pool in cmd line :

If set here, this will be the MAIN pool, you can also add more pools in pools.txt config file.

Parameters:

--cworker value (worker name or rig id - pool must support it)
--cpool url:port (pool address:port without stratum prefix)
--cwallet address (user wallet address)
--cpassword value (pool password)
--ctls value (use SSL/TLS, true or false)
--cnicehash value (force nicehash, true or false)


Examples:

Use 1 GPU with id 0 , intensity 120, 2 threads on algo cryptonight v7 on nanopool:

SRBMiner-CN.exe --ccryptonighttype normalv7 --cgpuid 0 --cgpuintensity 120 --cgputhreads 2 --cpool xmr-eu1.nanopool.org:14444 --cwallet 4A5hJyu2FvuM2azexYssHW2odrNCNWVqLLmzCowrA57xGJLNufXfzVgcMpAy3YWpzZSAPALhVH4Ed7x o6RZYyw2bUtbm12g.donation


Use 5 GPUS with id 0,1,2,3,4 , intensities 56,56,55,58,55, 2 threads for each GPU, on algo cryptonight v7 on nanopool:

SRBMiner-CN.exe --ccryptonighttype normalv7 --cgpuid 0,1,2,3,4 --cgpuintensity 56,56,55,58,55 --cgputhreads 2,2,2,2,2 --cpool xmr-eu1.nanopool.org:14444 --cwallet 4A5hJyu2FvuM2azexYssHW2odrNCNWVqLLmzCowrA57xGJLNufXfzVgcMpAy3YWpzZSAPALhVH4Ed7x o6RZYyw2bUtbm12g.donation



--------------------------------------------------------------------------
KEYBOARD SHORTCUTS
--------------------------------------------------------------------------


- Press 's' to see some basic stats
- Press 'h' to see hashing speed
- Press 'p' to fast switch to next pool from pools config file
- Press 'r' to reload pools, if you add a pool to pools config, no need to restart miner to use new pool
- Press number from 0-9 to disable/enable from gpu0-gpu9, then shift+0 for gpu10, shift+1 for gpu11..etc. until gpu19 max (use US keyboard where SHIFT+1 = !, SHIFT+2 = @ ..etc..)



--------------------------------------------------------------------------
CONFIGURATIONS
--------------------------------------------------------------------------
The X/X/X means : INTENSITY/WORKSIZE/THREADS

Vega56 (samsung and hynix) with stock and 64 bios

heavy/haven/bittubev2/italo : 56/8/2 ( ~1550+HR )
v7/stellitev7 : 112/16/2  or 120/8/2 ( ~2000+HR )
fast : 112/16/2 or 120/8/2 ( ~3800/3900HR )
lite/litev7 : 234/8/2 ( ~4200+HR )
alloy : 120/16/2 ( ~960/1050HR )
artocash : 112/16/2  ( ~1950HR )


Vega 64

heavy/haven/bittubev2/italo : 60/8/2 ( ~1550+HR )
v7/stellitev7 : 120/16/2 ( ~2100/2200+HR )
fast : 120/16/2 ( ~4000+HR )


Rx 550 2G

heavy/haven/bittubev2/italo : 14/12/2 ( ~350HR )
v7/stellitev7 : 27/12/2 ( ~520HR )
fast : 21/12/2 ( ~890HR )
lite/litev7 : 45/12/2 ( ~1050HR )

--------------------------------------------------------------------------
OVERDRIVENTOOL SETTINGS
--------------------------------------------------------------------------
The X/X means : MHZ/mV

Vega56 air ref (samsung) 64 bios
GPU: P7 = 1442/906
MEM: P3 = 1095/906


Vega56 (hynix) stock bios
GPU: P7 = 1448/930
MEM: P3 = 908/930


Vega 64 Liquid (samsung)
GPU: P7 = 1478/920
MEM: P3 = 1100/910


Rx 550 2G (elpida)
GPU: P7 = 1170/875
MEM: P3 = 1900/875

Rx 550 2G (hynix)
GPU: P7 = 1170/875
MEM: P3 = 2030/875


--------------------------------------------------------------------------
INFORMATION
--------------------------------------------------------------------------

You have to change the wallet address in sample config.txt file to your own, or you will donate some hashing power to me. (Thanks)
SRBMiner is using a heavily modified version of Wolf0/Wolf9466/OhGodAPet's OpenCL kernel.

Windows freeze/gpu crashing when we switch to another pools, i'm using miningrigrental to manage pools with 11 RX Vega.

[dragonmike]

Quick one: will the latest 18.7.1 driver slow down the miner if I try to mine Alloy? I think I installed it before the limitations on heavy were discovered. Normalv7 wasn't impacted then...

Just gauging whether I will have to downgrade my drivers (which I'd like to avoid if I can )!

18.8.1 are the latest, and idk try
On which drivers are you now?

I can't remember!
Been on holiday. Haven't turned on these rigs for three weeks...
Bah, it's not like I'm missing out on a lot of profits these days. Whenever I have some time, I'll fire them back up and check.

[hmgp]

Hi! First off, congrats on the work done on this miner.

I've been using cast and not happy with it and it became very unstable.

I've tried SRB using minimal config and it connected correctly. Immediately I saw better hashrate with the same configs, but it gets several jobs and sends no shares, which is really strange. So I started playing around with the settings, and now it doesn't even start  

These are my files. What am I doing wrong here? Also want to put the restart script in, but haven't tried yet.

I'd really appreciate if you could help me on getting SRB cracking =)

Thanks

PS: the GPU restart seems to do it one by one, then says it restarted correctly, and shuts down after a quick message in red

start file

setx GPU_MAX_HEAP_SIZE 100
setx GPU_MAX_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_MAX_SINGLE_ALLOC_PERCENT 100
@echo off
cd %~dp0

cd C:\Users\vega\Desktop\SRBMiner-CN-V1-6-5
OverdriveNTool.exe -p1Vega-h4 -p2Vega-h3 -p3Vega-h1 -p4Vega-h2
timeout /t 1

cls
set EXE=SRBMiner-CN.exe
%EXE% --sendallstales --config config.txt --pools pools.txt

reboot file

shutdown.exe /r /t 10

pools file

{
"pools" :
[
   {"pool" : "xmr-eu1.nanopool.org:14444", "wallet" : "XXXwallet.USER/email@email.com", "password" : "x"},
]
}

config file

{
"cryptonight_type" : "normalv7",
"intensity" : 0,
"double_threads" : true,
"restart_devices_on_startup" : true,
"min_rig_speed" : 7400,
"target_temperature" : 60,

/* THERE ARE A LOT OF OPTIONS/PARAMETERS YOU CAN USE HERE */
/* YOU CAN FIND THEM IN THE README FILE */
}

[ViWe]

Hi! First off, congrats on the work done on this miner.

I've been using cast and not happy with it and it became very unstable.

I've tried SRB using minimal config and it connected correctly. Immediately I saw better hashrate with the same configs, but it gets several jobs and sends no shares, which is really strange. So I started playing around with the settings, and now it doesn't even start  

These are my files. What am I doing wrong here? Also want to put the restart script in, but haven't tried yet.

I'd really appreciate if you could help me on getting SRB cracking =)

Thanks

PS: the GPU restart seems to do it one by one, then says it restarted correctly, and shuts down after a quick message in red

...

I only had GPU crashes with too aggressive intensity or too aggressive memstraps/overclock/undervolt settings.
So, you could try setting the intensity manually lower than SRBMiner now does it based on your setting 0. This might work, but more likely it is in the profile you use with Overdriventool. I don't have VEGa's, but general advice here would be to cut back what you are doing there more to stock value and then see if it improves. Once you have a stable situation, crank it up.
Hope this works for you. May be somebody else can give you his Vega settings as a reference to try.

[QuirkSilver]

Try with kernel 2 it will work

next version will set it automatically

That worked.  Thanks!

Did you use kernel 2 only for the r9 200 series or you had to use them on the R9 390 Series also?

edit: have you tried int: 30 , w: 8 , threads: 3 setup on your hawaii's ?

In my case for the GPUs R9 390 8GB - R9 390X 8GB (Hawaii Cards) always the best option is Threads: 3

I used to get best on non-claymore miners with 390 with 740-860 intensity/threads 2/work 4 that was with sgminer/ xmr stak/ gatelessgate old/new also, thats for CN normal 1040 mhz 830-840 h/s
older 290/290x go for Stilt's bioses blindly, i remember the last Hawaii i had (i burnt 4-5 of them ) was 90 watts chip power, doing 870 h/s-900 h/s.

[ikyo]

I noticed the issue again with the miner seeming to drop hashing speeds on cards by around 100 H/s per Vega 64 card.  I looked back through the mining log and didn't see anything strange.  I tried resetting the cards while the miner ran (Overdriventool -r and then -p) and that didn't fix the issue.  I closed the miner down and restarted it.  That caused all of the cards to go back up to where they were at originally. 

Code:

setx GPU_MAX_HEAP_SIZE 100
setx GPU_MAX_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_MAX_SINGLE_ALLOC_PERCENT 100

OverdriveNTool.exe -consoleonly -r1 -r2 -r3 -r4 -r5 -r6 -p1VEGA64 -p2VEGA64 -p3VEGA64 -p4VEGA64 -p5VEGA64 -p6VEGA64
timeout /t 5

SRBMiner-CN.exe --config monero-config.txt --pools monero-sxmr-pools.txt
pause

Code:

[Profile_0]
Name=VEGA64
GPU_P0=852;900;0
GPU_P1=991;900;0
GPU_P2=1084;900;0
GPU_P3=1138;900;0
GPU_P4=1150;900;0
GPU_P5=1202;900;0
GPU_P6=1212;905;0
GPU_P7=1408;925
Mem_P0=167;900;0
Mem_P1=500;900;0
Mem_P2=800;900;0
Mem_P3=1095;905
Fan_Min=3000
Fan_Max=4900
Fan_Target=75
Fan_Acoustic=2400
Power_Temp=85
Power_Target=0

[General]
MainWindowLeft=538
MainWindowTop=277

I just lowered it to 1095 from 1100 on the memory.

[hmgp]

I only had GPU crashes with too aggressive intensity or too aggressive memstraps/overclock/undervolt settings.
So, you could try setting the intensity manually lower than SRBMiner now does it based on your setting 0. This might work, but more likely it is in the profile you use with Overdriventool. I don't have VEGa's, but general advice here would be to cut back what you are doing there more to stock value and then see if it improves. Once you have a stable situation, crank it up.
Hope this works for you. May be somebody else can give you his Vega settings as a reference to try.
[/quote]


I don't think it's related to that. I know for experience I'm not pushing them hard and Vega's immediately complain when we taken them to the limit

Maybe it's because I'm using the latest drivers that it doesn't do any shares, and I'll revert to other drivers to test it, but I'd appreciate it if someone could help me fine tune the files to run the miner properly with the watchdog, restarting gpu's at the start, etc =)

[hmgp]

I noticed the issue again with the miner seeming to drop hashing speeds on cards by around 100 H/s per Vega 64 card.  I looked back through the mining log and didn't see anything strange.  I tried resetting the cards while the miner ran (Overdriventool -r and then -p) and that didn't fix the issue.  I closed the miner down and restarted it.  That caused all of the cards to go back up to where they were at originally. 

Code:

setx GPU_MAX_HEAP_SIZE 100
setx GPU_MAX_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_MAX_SINGLE_ALLOC_PERCENT 100

OverdriveNTool.exe -consoleonly -r1 -r2 -r3 -r4 -r5 -r6 -p1VEGA64 -p2VEGA64 -p3VEGA64 -p4VEGA64 -p5VEGA64 -p6VEGA64
timeout /t 5

SRBMiner-CN.exe --config monero-config.txt --pools monero-sxmr-pools.txt
pause

Code:

[Profile_0]
Name=VEGA64
GPU_P0=852;900;0
GPU_P1=991;900;0
GPU_P2=1084;900;0
GPU_P3=1138;900;0
GPU_P4=1150;900;0
GPU_P5=1202;900;0
GPU_P6=1212;905;0
GPU_P7=1408;925
Mem_P0=167;900;0
Mem_P1=500;900;0
Mem_P2=800;900;0
Mem_P3=1095;905
Fan_Min=3000
Fan_Max=4900
Fan_Target=75
Fan_Acoustic=2400
Power_Temp=85
Power_Target=0

[General]
MainWindowLeft=538
MainWindowTop=277

I just lowered it to 1095 from 1100 on the memory.

well, that's exactly what happens on my vega 56 rig and why I'm trying SRB (besides the fact I'm not overall happy with cast). Goes up just fine, let's say 3 vega do 6K. after a while, one drops, and it may be like that for hours, or really drop to ridiculously low hashrate, and/or more cards to the same.

lowering P7 or P3 does nothing to it... which drivers are you using?

[ikyo]

I noticed the issue again with the miner seeming to drop hashing speeds on cards by around 100 H/s per Vega 64 card.  I looked back through the mining log and didn't see anything strange.  I tried resetting the cards while the miner ran (Overdriventool -r and then -p) and that didn't fix the issue.  I closed the miner down and restarted it.  That caused all of the cards to go back up to where they were at originally. 

Code:

setx GPU_MAX_HEAP_SIZE 100
setx GPU_MAX_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_MAX_SINGLE_ALLOC_PERCENT 100

OverdriveNTool.exe -consoleonly -r1 -r2 -r3 -r4 -r5 -r6 -p1VEGA64 -p2VEGA64 -p3VEGA64 -p4VEGA64 -p5VEGA64 -p6VEGA64
timeout /t 5

SRBMiner-CN.exe --config monero-config.txt --pools monero-sxmr-pools.txt
pause

Code:

[Profile_0]
Name=VEGA64
GPU_P0=852;900;0
GPU_P1=991;900;0
GPU_P2=1084;900;0
GPU_P3=1138;900;0
GPU_P4=1150;900;0
GPU_P5=1202;900;0
GPU_P6=1212;905;0
GPU_P7=1408;925
Mem_P0=167;900;0
Mem_P1=500;900;0
Mem_P2=800;900;0
Mem_P3=1095;905
Fan_Min=3000
Fan_Max=4900
Fan_Target=75
Fan_Acoustic=2400
Power_Temp=85
Power_Target=0

[General]
MainWindowLeft=538
MainWindowTop=277

I just lowered it to 1095 from 1100 on the memory.

well, that's exactly what happens on my vega 56 rig and why I'm trying SRB (besides the fact I'm not overall happy with cast). Goes up just fine, let's say 3 vega do 6K. after a while, one drops, and it may be like that for hours, or really drop to ridiculously low hashrate, and/or more cards to the same.

lowering P7 or P3 does nothing to it... which drivers are you using?

I am using 18.5.1 as suggested in the original post.  They don't drop to nothing.  They just go down.  Once I restart the miner, everything is fine.  It is almost like a memory leak or something.  I have had it happen on 2 different systems with different cards so we shall see.

[uzgidebas]

Are you planning to release linux version?