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

[doktor83]

Hey Doktor

I was using the miner with adrenalin 17.x.1 version which month i don't remember exactly and ADL wasn't working. I upgraded the driver with adrenalin 18.5.1 and ADL problem is solved. Maybe you can add this as known issues. Keep up with good work. Regards.

i think you can use it even on 17.x drivers since i added support for Overdrive5. But i did not test this lol 

[doktor83]

how do i properly use this command? "restart_devices_on_startup_script" :

i inserted in the config.txt like this

"restart_devices_on_startup_script" : Run_OverdriveNTool.bat,

miner goes to start then kills itself.

also anyway to control enable/ disable on which Vega's. Not sure if its the card or the riser, but one of my Vega's cant handle the disable/enable with out crashing the whole OS.

TIA

Anyone?

"restart_devices_on_startup" : IF TRUE IT WILL USE DEVCON TO DISABLE/ENABLE EVERY VEGA GPU IN YOUR MACHINE BEFORE MINING STARTS
"restart_devices_on_startup_script" : FILENAME, THIS SCRIPT RUNS AFTER VEGA GPU ENABLE/DISABLE PROCESS, GOOD FOR SETTING UP OVERCLOCKING

so first you have to set "restart_devices_on_startup" : true, then "restart_devices_on_startup_script" : Run_OverdriveNTool.bat, where Run_OverdriveNTool.bat is in miner dir.

You can disable/enable every vega at once, no option to do this card by card.

Thank you Sir! Love your Miner! learned about it from Geek Mark, switched all my rigs to it! also getting my friends to switch too! thank you for the assistance an continued development of the software!

Thanks
Unfortunately i dont have a Vega card, but lot of folks here do, so im sure if you have any vega specific questions, they will help you!

[argai666]

Hey Doktor

I was using the miner with adrenalin 17.x.1 version which month i don't remember exactly and ADL wasn't working. I upgraded the driver with adrenalin 18.5.1 and ADL problem is solved. Maybe you can add this as known issues. Keep up with good work. Regards.

i think you can use it even on 17.x drivers since i added support for Overdrive5. But i did not test this lol 

I did not tried that one too I am using OverdriveNtool btw.

[doktor83]

When I use the function enable and disable vega card, I get the following error: cl_out_of_host_memory when creating clCreateCommandQueue for DeviceID 0 (Thread 0) , does somebody has any idea?

Make sure you have this in start.bat at top :

setx GPU_FORCE_64BIT_PTR 1
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


Also try other video driver, for ex. 18.3.4

[doktor83]

Hey Doktor

I was using the miner with adrenalin 17.x.1 version which month i don't remember exactly and ADL wasn't working. I upgraded the driver with adrenalin 18.5.1 and ADL problem is solved. Maybe you can add this as known issues. Keep up with good work. Regards.

i think you can use it even on 17.x drivers since i added support for Overdrive5. But i did not test this lol 

I did not tried that one too I am using OverdriveNtool btw.

that's ok , its a great tool to set your clocks and undervolt.

Version of overdrive can be set in gpu_conf for every card separately :

"adl_type" :
1-OverdriveN
2-Overdrive5



"gpu_conf" :
[
   { "id" : 0, "intensity" : 0, "worksize" : 8, "threads" : 2, "adl_type" : 1},
   { "id" : 1, "intensity" : 0, "worksize" : 8, "threads" : 2, "adl_type" : 2}
]

[doktor83]

On AMD 18.5.1 driver on 1803 Windows 10 x64 my RX 580 recognized by SRB miner as R9 200 series [Ellesmere] and every launch build new kernels... It's abnormal. I think, you must update miner to support new drivers, doctor...
By the word, all Claymore's miners detects my cards right on 18.5.1...

And I still can't understand, why every lounch of SRB miner speed is different from each other... To reach max speed it's need to launch SRB several times without any changes in config or something else...
Cards is RX 580 8Gb.

Are you mining Heavy?  If that's the case then the behavior is existent on multiple cards.

Yes, heavy. But case is not in cars. Case is in miner.

I get similar behavior on xmr-stak on Heavy.  

xmr stak, gateless gate also have this 'problem'.
Maybe this means the 'problem' is in the algo, not in miners?

What about compiling kernels every launching?

possibly something to do with read/write permissions for the miner, cause video driver has nothing to do with this, and 1804 v is working fine as someone wrote.
And yeah, if 18.5.1 recognises RX580 as R9 then its a really f**ked up driver, something really buggy there.
Yes claymore shows ok probably sees "Ellesmere XT" ah ok that is 480/580 lol

[O-Coin]

GPU_CONF Help

Trying to set double threads on 8-10 of 12 RX580  I cant set all 12 due to not enough Vmem I already have an 8XRX580 rig running on double threads and Avg 8.12 khs.  12 card running only single thread only getting 9.6khs so I want to run as many as i can on double thread to maximize hash.  A bigger ssd would work also but only have the 120gb and have v mem at 72gb
I am using Awesome miner
I havent been able to get any GPU_conf to work only intensity and doublethreads true or false soon as I try setting each card miner dont start.  Any help much appreciated.

{
  "cryptonight_type": "heavy",
  "intensity": 63,
  "double_threads": false,
  "api_enabled": true,
  "api_port": 4028
"gpu_conf" :
[
{ "id" : 0, "intensity" : 62, "double_threads" : false},
{ "id" : 1, "intensity" : 62, "double_threads" : false},
{ "id" : 2, "intensity" : 62, "double_threads" : false},
{ "id" : 3, "intensity" : 0, "double_threads" : true},
{ "id" : 4, "intensity" : 0, "double_threads" : true},
{ "id" : 5, "intensity" : 0, "double_threads" : true},
{ "id" : 6, "intensity" : 0, "double_threads" : true},
{ "id" : 7, "intensity" : 0, "double_threads" : true},
{ "id" : 8, "intensity" : 0, "double_threads" : true},
{ "id" : 9, "intensity" : 0, "double_threads" : true},
{ "id" : 10, "intensity" : 0, "double_threads" : true},
{ "id" : 11, "intensity" : 0, "double_threads" : true}
]
}


Thanks

[hesido]

GPU_CONF Help

Trying to set double threads on 8-10 of 12 RX580  I cant set all 12 due to not enough Vmem I already have an 8XRX580 rig running on double threads and Avg 8.12 khs.  12 card running only single thread only getting 9.6khs so I want to run as many as i can on double thread to maximize hash.  A bigger ssd would work also but only have the 120gb and have v mem at 72gb
I am using Awesome miner
I havent been able to get any GPU_conf to work only intensity and doublethreads true or false soon as I try setting each card miner dont start.  Any help much appreciated.

{
  "cryptonight_type": "heavy",
  "intensity": 63,
  "double_threads": false,
  "api_enabled": true,
  "api_port": 4028
"gpu_conf" :
[
{ "id" : 0, "intensity" : 62, "double_threads" : false},
{ "id" : 1, "intensity" : 62, "double_threads" : false},
{ "id" : 2, "intensity" : 62, "double_threads" : false},
{ "id" : 3, "intensity" : 0, "double_threads" : true},
{ "id" : 4, "intensity" : 0, "double_threads" : true},
{ "id" : 5, "intensity" : 0, "double_threads" : true},
{ "id" : 6, "intensity" : 0, "double_threads" : true},
{ "id" : 7, "intensity" : 0, "double_threads" : true},
{ "id" : 8, "intensity" : 0, "double_threads" : true},
{ "id" : 9, "intensity" : 0, "double_threads" : true},
{ "id" : 10, "intensity" : 0, "double_threads" : true},
{ "id" : 11, "intensity" : 0, "double_threads" : true}
]
}


Thanks

Funny thing.. I have 13 cards and I was at the Virtual Mem limit, so I bought *today* an HDD (sdd / hdd does not matter in performance for vmem used by miners)..

So I immediately jumped at 60 intensity which everybody were happily using.. Miner crashed..
Set down to 58, some gpu's don't work..
Set down to 54 (restarting between each setting as the miner cannot start once you have a crashed gpu due to intensity..) and volia, it worked, adding a respectable 400mh/s total over the 51 intensity setting (which was set by auto), and cards running using a about 7100 MB memory, not shabby, out of 8MB that's good.. And 95MB vmem was used, so this was not possible with my SSD.

I restarted miner a couple of times and it worked.

So I decided to push for 55.. Sheit. Problem with a few gpu's again... Without hesitation, downloaded the latest SRB, 1.5.6, maybe it fixed something.. It did not work either.

But this time, I can't do 54, which I was able to before!  This, no matter how many restarts and clean boots I did :\

Went back to 52. 53 even did not work. I bought a hard disk for a +1 intensity change :\ (Using double threads though..)

I too would like to hear from you people...

Edit: Back to auto setting, 51 :\ 51 seems to be the only stable setting for me and that's when I set it to auto. (It somehow is problematic when I set 51 manually! Why could that be... That's very odd, maybe the auto setting tingles some parts that prevent the GPU from crashing in mem-allocation???)

[adaseb]

Anyone here still mining with the good old Radeon 7970 and R9 280X GPUs.

Wondering what the most optimized intensity would be to set. So far using 2 thread causes entire system to freeze up. But looks like the intensity and work size could be adjusted.

With Claymore I get like 600hs with the Tahitis wondering if I can get more speed with this software?

I tried on 7950, CN Heavy (XHV) = 300 h/s

Yes I got more or less the same whether its a 7970 or a 280X. Pretty much Pitcairns also got a similiar speed.

I tried increasing the threads and got an instant freeze and also tried to increase the intensity and got a speed of 0H/s.

Its too bad because CN on the old Tahitis is very power efficient and better than mining ZEC even if the profits aren't as good.

[henri2018]

On AMD 18.5.1 driver on 1803 Windows 10 x64 my RX 580 recognized by SRB miner as R9 200 series [Ellesmere] and every launch build new kernels... It's abnormal. I think, you must update miner to support new drivers, doctor...
By the word, all Claymore's miners detects my cards right on 18.5.1...

And I still can't understand, why every lounch of SRB miner speed is different from each other... To reach max speed it's need to launch SRB several times without any changes in config or something else...
Cards is RX 580 8Gb.

Are you mining Heavy?  If that's the case then the behavior is existent on multiple cards.

Yes, heavy. But case is not in cars. Case is in miner.

I get similar behavior on xmr-stak on Heavy.  

xmr stak, gateless gate also have this 'problem'.
Maybe this means the 'problem' is in the algo, not in miners?

What about compiling kernels every launching?

possibly something to do with read/write permissions for the miner, cause video driver has nothing to do with this, and 1804 v is working fine as someone wrote.
And yeah, if 18.5.1 recognises RX580 as R9 then its a really f**ked up driver, something really buggy there.
Yes claymore shows ok probably sees "Ellesmere XT" ah ok that is 480/580 lol

I use latest driver 18.5.1 on my 6 Vega56 rig and SRB 1.5.6. No issue at all. Vega is recognized as RX Vega (GFX900), no different with previous SRB version and driver version.

[knittycatkitty]

When I use the function enable and disable vega card, I get the following error: cl_out_of_host_memory when creating clCreateCommandQueue for DeviceID 0 (Thread 0) , does somebody has any idea?

usually it is virtual memory allocation for me when i get that error

[duyquang06]

Auto intensity is good now, leave all rigs to intensity:0, no crash at all.
Really good miner, Doc, +1 for u.

[hesido]

Dok, how do you figure auto-intensity?

Because it seems to be so precise, it sets 51 for me and even manual setting with 52 I run into big problems :\

[amorphis22]

Hello Doc!

Next week XTL(stellite coin) be forked to new algo CryptoNight variant xtl.

Do you plan to add this algo to SRB miner?
It'll be very useful feature.

P.S. xmrig have it already

WBR

[livada]

Dok, how do you figure auto-intensity?

Because it seems to be so precise, it sets 51 for me and even manual setting with 52 I run into big problems :\

intensyity = 0

this is auto

[Larvitar]

Hello Doc!

Next week XTL(stellite coin) be forked to new algo CryptoNight variant xtl.

Do you plan to add this algo to SRB miner?
It'll be very useful feature.

P.S. xmrig have it already

WBR

Ya. Is a CN-V7 version with minor changes. https://github.com/stellitecoin/xtl-stak/blob/0be955abb58aa8ce9e24283aa16f6753c7a00b55/xmrstak/backend/cryptonight.hpp

[doktor83]

V1.5.7
- Added support for StelliteV4 algo
- Added option "worker" in pool config - pool must support this, also CMD parameter --cworker
- Fixed average hashrate display (1,5,30 min) when there are disabled GPU's

+ Had to push this quick update because of the upcoming Stellite fork, so you guys can get ready on time
+ "worker" option won't work on every pool, because pool must support 'rigid' parameter

It would look something like this :

{
"pools" :
[
   {"pool" : "xmr-eu1.nanopool.org:14444", "wallet" : "your-wallet", "password" : "x", "worker" : "my-rig-name"}
]
}

[balrogbr]

Hello, is there any operating system type, ethOS or hiveOS that supports this SRB miner? I would love to use it however use ethOS!

[Zazax21]

I've downloaded the V1.5.7 and connected it to etn.nanopool, algo monero7, it seems that it does not give share, only recieve job from pool, can someone check it ? i've tried on cn-heavy algo and it's working.

[doktor83]

I've downloaded the V1.5.7 and connected it to etn.nanopool, algo monero7, it seems that it does not give share, only recieve job from pool, can someone check it ? i've tried on cn-heavy algo and it's working.

diff at nanopool is 240009, also the fork was activated few blocks ago, and its normalv7 here, not monero7.

Looks like i screwed something when implementing Stellite, please don't yet download 1.5.7, i will fix it ASAP an re-upload.

All fixed