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

[Marvell2]

final issue I have is super low hashrate on vega frontier 16gb cards
using the default settings.

[GKumaran]

Dok, can u give us a quick tip as to what it is and how to tune "bralock" parameter and recommended values for Vega 64. The readme just says 1-256.

[seefatlow]

Dok, any possibility of including GPU manufacturer's name in the GPU list instead of just GPU# for easier troubleshooting?

[alexgsx]

1 rig  4 x vega56 + 2 x rx570   9400hs  1040watts at the wall

All my rx 570s hashing exactly same on v8 940hs.
My rx 550s are at 450hs each

which setting do you have for your rx570 and biosmod?

[11th_miner]

Hi, so far srbminer is the best miner for v8 mining; doktor thanks for that.
I had a lot of issues to get the best set up for my vega64 rigs and v8, and I had to remove a card from my rigs to fix the power problem( my rig had five vega64 and 1200watt cooler master, now it's a four card rig  ), thanks to intensity set up from doktor that's my result
https://ibb.co/fEZGt0
, for me the main problem is the HBM temperature, anyone had a better temperature for this hash rate?

[RuMiner]

Has the Claymore monitor tool for SRB miner been fixed yet ?  I think the newest SRB release broke it

ClayConn works ok. is there another tool?
PS API settings moved from config to start string, like SRBMiner-CN.exe --config config.txt --pools pools.txt --logfile logs\%date%-%random%.log --apienable --apiport 21555 --apirigname rig10

[livada]

1.6.8 and 1.6.4 +18.10.1 driver not work

not valid share on any algo.(try v8-v7-fast-arto-xtl)

Complie with 18.10.1. and try with precompile binnayr  but not work.

[Amigo9881]

dok, my Vega FE edition is not able to push beyond  ~1300 h/s .

i even let the card compile kernels again.

Can anyone suggest solution ?

Thanks

[Marvell2]

Has the Claymore monitor tool for SRB miner been fixed yet ?  I think the newest SRB release broke it

ClayConn works ok. is there another tool?
PS API settings moved from config to start string, like SRBMiner-CN.exe --config config.txt --pools pools.txt --logfile logs\%date%-%random%.log --apienable --apiport 21555 --apirigname rig10

clayconn wasn’t working for me, I guess i was using the wrong config files , will try again tomorrow

[Marvell2]

dok, my Vega FE edition is not able to push beyond  ~1300 h/s .

i even let the card compile kernels again.

Can anyone suggest solution ?

Thanks

same issue here, It might be a bug c8 implementations? I see even worse results with cast xmr
I tried rasing the intesity to 140 and i got up to 1500 but , something is really off with these frontiers

[playfast]

I have a rx560 rig using 18.6.1 drivers, on v8 they all went down to ~380-400 h/s with the same settings from v7 (500-520 h/s)
I had power limit at -10% with v7, on v8 I can get back to ~430-460 h/s just by raising the power limit to +10%
Some cards got a small boost when I increased virtual memory from 16 gb to 20 gb. Haven't tried to change the intensity and worksize yet, doubt they will make any difference. v8 definitely needs more power and virtual memory than v7 to alleviate hash loss.

[coke15]

hi

i was virulent few month ago, with the "lost" hasrhate between rig and pool.

Now with the 1.6.8 ,the hashrate is good.

nice to see you play game,thanks

[Lucky - Luciano]

Hi, you can use SRBminer: https://bitcointalk.org/index.php?topic=3167363.0

Low fee, great speed, 950h on my RX580s and supports v8 fork.

Damn, I have only  836 H/s. What settings do you use? These are my settings:

{
"cryptonight_type" : "normalv8",
"intensity" : 96,
"worksize" : 8,
"double_threads" : true
}

[Bojcha]

Hi, you can use SRBminer: https://bitcointalk.org/index.php?topic=3167363.0

Low fee, great speed, 950h on my RX580s and supports v8 fork.

Damn, I have only  836 H/s. What settings do you use? These are my settings:

{
"cryptonight_type" : "normalv8",
"intensity" : 96,
"worksize" : 8,
"double_threads" : true
}

intens 58 and worksize 16

[camchimonline]

How to disable 1 gpu, dont use press 0-9 key
Thank

[logicus]

I am using latest version 1.6.8  and ati drivers 18.6.1. Mining rig is 4 x Vega 56 (flashed with Vega 64 bios).
Core clock 1450@900mv memory clock 1100@900mv . /Hellae's powerplay table mod/
v7 - 8200 h/s @ 700 W at the wall.
v8 - 7800 h/s @ 800 W at the wall.

config_normalv8.txt
{
"cryptonight_type" : "normalv8",
"intensity" : 0,
"double_threads" : true,
"target_temperature" : 55,
"shutdown_temperature" : 80,
"persistent_memory" : true,
/*"gpu_conf" :
[
   { "id" : 0, "intensity" : 120, "worksize" : 16, "threads" : 2},
   { "id" : 1, "intensity" : 120, "worksize" : 16, "threads" : 2},
   { "id" : 3, "intensity" : 120, "worksize" : 16, "threads" : 2},
   { "id" : 4, "intensity" : 120, "worksize" : 16, "threads" : 2}
]*/
}

Check https://docs.google.com/spreadsheets/d/1pH326fwnYWC5Jef2FdF4KnJ4CLj0cwDDtGjrlgekk1A/edit#gid=514064850

[coke15]

I have 5 vega56 Adrenalin 18.6.1
3 gfx 900 give by 1850 ~
2 gfx 901 give 1000 ~
what is the problem?

why my vega 901 give 1000 h/s
Vega 900 1800 h/s
How to raise hashrate on 901?

If you are using the binary kernels try to rename that folder and let the miner compile it again.

hi same here:
not working for me:
vega 901 give 1200
vega 900 give 1200
?

[tvukoman]

I am using latest version 1.6.8  and ati drivers 18.6.1. Mining rig is 4 x Vega 56 (flashed with Vega 64 bios).
Core clock 1450@900mv memory clock 1100@900mv . /Hellae's powerplay table mod/
v7 - 8200 h/s @ 700 W at the wall.
v8 - 7800 h/s @ 800 W at the wall.

config_normalv8.txt
{
"cryptonight_type" : "normalv8",
"intensity" : 0,
"double_threads" : true,
"target_temperature" : 55,
"shutdown_temperature" : 80,
"persistent_memory" : true,
/*"gpu_conf" :
[
   { "id" : 0, "intensity" : 120, "worksize" : 16, "threads" : 2},
   { "id" : 1, "intensity" : 120, "worksize" : 16, "threads" : 2},
   { "id" : 3, "intensity" : 120, "worksize" : 16, "threads" : 2},
   { "id" : 4, "intensity" : 120, "worksize" : 16, "threads" : 2}
]*/
}

Check https://docs.google.com/spreadsheets/d/1pH326fwnYWC5Jef2FdF4KnJ4CLj0cwDDtGjrlgekk1A/edit#gid=514064850

logicus that is great result  

Please tell me, what cooling you have for rig, what is the temperature core and HBM of the cards, and RPM?

[Int3ll]

Hello @all, i have 2 rigs with 7x RX550 each ( Elpida 2GB RAM ) with Windows 10 and SRBMiner 1.6.8 on both.
One is using Adrenaline 18.7.1 and hashes 2.8K ( down from 3.4 K on v7 ) and one is using Adrenaline 18.5.1 and hashes 2.5K ( down from 3.2 ) but did not find any share for the past 4 hours.

Is there a problem with 18.5.1 on Win 10 and RX550?  

RIG 1 => 2.5k ( intensity 26, worksize 24, theads 2) ( RX550 with 8 Compute units )  using Adrenaline 18.5.1 but no shares are being found
RIG 2 => 2.8K ( intensity 26, worksize 24, theads 2) ( RX550 with 10 Compute units ) using Adrenaline 18.7.1

P.S. If i set the worksize to above 24 ( ex:26 or 32) it BSOD's instantly. Maybe i overclocked them to hard???   ( 1000 core and 1950 Memory, without GPU errors on HWinfo ) and worked great on v7.
Power limit is set to +20%, because i read a few pages back that v8 uses more power.

[Beoga]

Does this miner work with AMD Radeon R9 200 Series?
I have problem with:
......
Initializing OpenCL devices, please wait..

GPU0: AMD Radeon R9 200 Series [pitcairn] [2048 MB][Intensity 26.0][W: 16][T: 2]
[K: 1][BUS: 1]

---------------------------------------------------------
SRBMiner Cryptonight AMD GPU miner 1.6.8
DevFee ~0.85%
Press 's' to get some stats
Press 'h' to get some hashrate
Press 'r' to reload pools
Press 'p' to switch to the next pool
Press 'o' to switch to the previous pool
Press 0-9 to disable/enable GPU while mining
Give it a few minutes to reach maximum hashrate
---------------------------------------------------------
Please use : Adrenalin 18.5.1, 18.5.2 or 18.6.1 drivers
Total pools: 2

Please wait...

CryptonightV8 mode enabled

[2018-10-20 21:20:31] Heating up system, please wait...
[2018-10-20 21:20:31] Compiling kernel for DEVICE BUS ID 1 ...
WARNING: Linking two modules of different data layouts!
WARNING: Linking two modules of different data layouts!
LLVM ERROR: Cannot select: 0x999d0d0: i32 = setcc 0x9997470, 0x5bdbb10, 0x99a2e2
0 [ORD=3092] [ID=246]
               .............