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

[_ap_]

Hi!

I have some info to the doc, and I have some questions First the infos:

My cards:
Sapphire Nitro+ rx 470 4gb hynix/elpida, with otc patch. Core 1300/1050, mem 2000/850 powlim -10% (afterburner). SRB 1.6.7, intensity: 0 (29), ws:8, th: 2.
Sapphire Nitro+ rx 580 4gb hynix/elpida, with otc patch. Core 1340/1000, mem 2000/850 powlim -10% (afterburner). SRB 1.6.7, intensity: 0 (29), ws:8, th: 2.
Asus dual oc rx 580 8gb hynix, with otc path. Core 1250/800 mem 2200/1000 powlim 0% (afterburner). Cast XMR 1.40, intensity 10.

My hashrates
SN+ 470: 740 H/s
SN+ 580: 760H/s
Asus 580: 920H/s

Watt summary: 295W (wattmeter)
Adrenalin driver: 18.5.1
Coin: BitTube

Second, my questions:
The Sapphire cards have higher hashrates with SRB (+ 40-50 H/s). The Asus is better with Cast XMR (+ 60-80 H/s). Why?
Which is the best adrenalin driver for BitTube?
I think the Sapphire cards hashrate isn't the best. Any idea how can I increase?

Thanks!

[1675447606]

1675447606

[1675447606]

1675447606

[1675447606]

1675447606

[Insuromancer]

I was mining at around 800 with beta blockchain driver. 1xRX480. No afterburner.
Now hashing around 300 after changing to recommended driver and "setting compute mode". I used DDU. All parameters are equal.
Can someone help?

you really did not set to compute mode

What did I do wrong? What should I do?

http://srbminer.com/start.html

Look at point nr 3.

PowerShell requested I use "./SRBMiner-CN.exe --SetComputeMode" as the command. The first one didn't take as you noted. Thank you, this resolved it!

[doktor83]

The Asus is better with Cast XMR (+ 60-80 H/s). Why?
Which is the best adrenalin driver for BitTube?

Try to set manually 56/8/2 for the 580 8g card.
Adrenalin 18.5.1 or 18.6.1 is recommended, if you want to do the 'warmup' stuff use 18.3.4 and maybe you can get out more hash on bittube algo.

[avvis]

i am looking for suggestions on how to solve the error CL_Build_program_failure.   I have received this error since the 1.6 version upgrade and as a result continue to run the 1.5.3 version.   I would like to upgrade but have not been able to get around this issue.

appreciate suggestions.

what gpu? some very old gpu?

GPU is the AMD R7 model.     Still working under the 1.5 version of SRB and with other miners such as stak and xmrig.

Try kernel 2, you can see on the first page how to set up gpu_conf and use kernel 2

Is your card Oland, Cape Verde, Bonaire or Pitcairn?

I'll give that a try,  my card is oland.

I tried different configurations with the Kernel settings and received the same results each time.

[livada]

Update intensity now
 changed:
on vegas  from 60 to 61, now 1520 h
on 580s from 54 to 58, now 1210 h

i post here before 20 day HR with  vega 56/64 referent card and HR is 1600+ on TUBE

Why you use 60/61 and have only 1520? 56 int is best for tube -haven-heavy

my post is here
https://bitcointalk.org/index.php?topic=3167363.msg45040757#msg45040757

61/9/2 , probaj

probao i odistao nakon sat vremena probavanja
to radi jedino ak nisi kartice vec natjero na krajnje granice.
kad kartica daje na 56 int 1600+HR onda sa 61/9 pukne nakon 15 sekundi a da sad krecem mjenjat napone i sppustat frekve mi se i neda jer sto cu onda dobit?

Ak imas vise od 1600Hr na tube sa 900mv mem i gpui ak 4 kartice od toga jedna LE koja trosi 30W vise jer je LE i trosi 700W iz zida onda cu probat opet? ak to sve imas onda cu se poigrat al za 5 HR da rusim sve i jebem se opet danima bas mi i nije neki djir. ktome nitko ovdje nije stavio neki SS da to fercera. stavio si ti kao 1500+HR sto je daleko ispod mojeg HR pa mi to nis ne znaci. ja sam opet na 1834 driverima i jedino sa njima i v7 algoom mi radi ta fora. sa nijednim drugim algoom to ne radi kod mene i to samo na referentnim samsung karitcama dok na nitru s hynixom to nicem ne sluzi ni u v7 algu. lili daje manji HR ili bloka kartice.

jedino  na referentnima na sTL/v7 radi s ovim postavkama
{ "id" : 0, "intensity" : 120, "worksize" : 16, "threads" : 1},
{ "id" : 0, "intensity" : 122, "worksize" : 16, "threads" : 1},
{ "id" : 1, "intensity" : 120, "worksize" : 16, "threads" : 1},
{ "id" : 1, "intensity" : 122, "worksize" : 16, "threads" : 1},
{ "id" : 2, "intensity" : 120, "worksize" : 8, "threads" : 1},
{ "id" : 2, "intensity" : 122, "worksize" : 8, "threads" : 1},
{ "id" : 3, "intensity" : 120, "worksize" : 16, "threads" : 1},
{ "id" : 3, "intensity" : 122, "worksize" : 16, "threads" : 1},

vege 56 sam isprobo w od 4 do 20 i najbolje daje na 16
vega 64 na w=9 daje osjetno manji hr

[SmartOdissey]

must disable/enable and first start with 60int 1 min and start again with 56 int.

how you do that automatically?

[avvis]

i am looking for suggestions on how to solve the error CL_Build_program_failure.   I have received this error since the 1.6 version upgrade and as a result continue to run the 1.5.3 version.   I would like to upgrade but have not been able to get around this issue.

appreciate suggestions.

what gpu? some very old gpu?

GPU is the AMD R7 model.     Still working under the 1.5 version of SRB and with other miners such as stak and xmrig.

Try kernel 2, you can see on the first page how to set up gpu_conf and use kernel 2

Is your card Oland, Cape Verde, Bonaire or Pitcairn?

I'll give that a try,  my card is oland.

I tried different configurations with the Kernel settings and received the same results each time.

that change did work - just didnt configure it properly -- Thank You!!!

[_ap_]

The Asus is better with Cast XMR (+ 60-80 H/s). Why?
Which is the best adrenalin driver for BitTube?

Try to set manually 56/8/2 for the 580 8g card.
Adrenalin 18.5.1 or 18.6.1 is recommended, if you want to do the 'warmup' stuff use 18.3.4 and maybe you can get out more hash on bittube algo.

The Asus 580 get lower hashrate with these values, I tried already.
What means 'warmup' stuff? More heat on cards?

[Turkish88]

Very bad expirience i have today.
Miner stopped mining with messege can't connect to the dev pool, you are blocked
Hey its not my problem if your dev pools not working, maybe you review your policy

[doktor83]

Very bad expirience i have today.
Miner stopped mining with messege can't connect to the dev pool, you are blocked
Hey its not my problem if your dev pools not working, maybe you review your policy

You should check your firewall or antivirus apps policy. This can happen only if it can't connect to a devfee pool minimum 3 or more times in a row (and when it can't connect to a devfee pool for the first time, it then tries another devfee pool when devfee mining time comes again, so the chances that a few different pools aren't available are i would say impossible).

So in practice it couldnt connect to different pools in a timespan of minimum few hours. Well you must admit that looks like something is blocking the outgoing miner connections.

[hesido]

Hello, DOK! Could you turn back bc driver support. The speed of my rx480 is the same compering with adrenalin driver but less power consumption. Now on algo switch my miner freezes(

You can use miner with BC drivers, who said you can't

Thanks for reply. But, in this case I don't know why miner sometime freezes on switch algo (moneroocean)

Describe it a little bit, or maybe a screenshot, log ?
Does the whole rig freeze, or only SRBMiner ? i need more info

Similarly on my end, on algo change, just at the warmup, it may free on algo change, and this is not picked up by the watchdog, the rig does not freeze, just SRBminer, with a probably crashed gpu. The problem is that the watchdog does not pick this at the start. Manually re-applying OverdriveNTool and restarting solves the problem. I had earlier sent the log here.

[doktor83]

Hello, DOK! Could you turn back bc driver support. The speed of my rx480 is the same compering with adrenalin driver but less power consumption. Now on algo switch my miner freezes(

You can use miner with BC drivers, who said you can't

Thanks for reply. But, in this case I don't know why miner sometime freezes on switch algo (moneroocean)

Describe it a little bit, or maybe a screenshot, log ?
Does the whole rig freeze, or only SRBMiner ? i need more info

Similarly on my end, on algo change, just at the warmup, it may free on algo change, and this is not picked up by the watchdog, the rig does not freeze, just SRBminer, with a probably crashed gpu. The problem is that the watchdog does not pick this at the start. Manually re-applying OverdriveNTool and restarting solves the problem. I had earlier sent the log here.

The watchdog is for initialised and running gpu's, so at the phase of opencl platform etc detection it is not running yet.

[logicalray77]

Hi everyone, I`m new to SRBMiner.
Can someone please share a working Nicehash config please.
I tried but with no luck.

[TeCe]

Next version in a day or two

have a litle trouble

algo heavy

cards rx470-570 4gb
double treads work fine
in config "intensity" : 29, "double_threads" : true

single treads dont work with error
Error CL_INVALID_BUFFER_SIZE when creating scratchpad buffer for DeviceID 0 (Thread 0)
in config "intensity" : 57, "double_threads" : false

if set intensity 0 setting 44 and slow rate
same rig work fine with config "intensity" : 58, "double_threads" : false
same drivers version 18.2.1 same swap size 60gb for 12 card same windows version 1709

CL_INVALID_BUFFER_SIZE errors means you're running out of GPU mem.  Reduce your intensity.

do you read my message?
rx470 4gb card
single treads
max Intensity 45
at Intensity 46 have error

Greetings!

Tell me, did you solve the issue with an intensity of 45 in the srb miner ... encountered the same thing ...

broke the Windows account, then created a new one, but the miner stopped running with an intensity higher than 45 ...

Removing drivers, reinstalling all the software, does not help yet!

Such a moment is observed only in Heavy, on other aglorhrams the intensity is set out to be necessary !!! !!!

[tg88]

Hi everyone, I`m new to SRBMiner.
Can someone please share a working Nicehash config please.
I tried but with no luck.

edit the pools.txt:
Example:

{"pool" : "stratum+tcp://cryptonightheavy.br.nicehash.com:3364", "wallet" : "xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx.worker_name", "password" : "x"},

[Coinseeker18]

Hi, I,m using your excelent miner doc but i have some inestability with my vegas, i have 5 vegas in a rig 2 vega 64 1 vega 56 @vega64 bios and 2 vega 56, my windows got me bsod with video scheduler internal error, and i answer if i could use the block chain driver instead the adrenaline 18.6.1? PD FYC Lately the vega 56 with bios 64 hang up with black screen and fan go to max rev. And i have to reset the rig. what do you suggest?

[hesido]

Hello, DOK! Could you turn back bc driver support. The speed of my rx480 is the same compering with adrenalin driver but less power consumption. Now on algo switch my miner freezes(

You can use miner with BC drivers, who said you can't

Thanks for reply. But, in this case I don't know why miner sometime freezes on switch algo (moneroocean)

Describe it a little bit, or maybe a screenshot, log ?
Does the whole rig freeze, or only SRBMiner ? i need more info

Similarly on my end, on algo change, just at the warmup, it may free on algo change, and this is not picked up by the watchdog, the rig does not freeze, just SRBminer, with a probably crashed gpu. The problem is that the watchdog does not pick this at the start. Manually re-applying OverdriveNTool and restarting solves the problem. I had earlier sent the log here.

The watchdog is for initialised and running gpu's, so at the phase of opencl platform etc detection it is not running yet.

That's what I thought. There are some nice things that can be done about this that everybody would appreciate It involves connecting early to pool, starting watchdog early, starting gpu's as they load the opencl code. The benefits would be four-fold, the hashing would start slightly earlier, watchdog would be able to catch a failed init, no time would be wasted if the active algo is different at the initial run, and there would be less strain on the PSU's that supply power during the ramp up. The last one would be very nice on the PSU's and would add more stability for the initializing sequence.

I know you are extremely busy with the monero forking and all but I hope you could take a look at this option!

edit: Have you even tried my proxy yet? It works quite nice with your miner

[Amgine]

Hi, I,m using your excelent miner doc but i have some inestability with my vegas, i have 5 vegas in a rig 2 vega 64 1 vega 56 @vega64 bios and 2 vega 56, my windows got me bsod with video scheduler internal error, and i answer if i could use the block chain driver instead the adrenaline 18.6.1? PD FYC Lately the vega 56 with bios 64 hang up with black screen and fan go to max rev. And i have to reset the rig. what do you suggest?

I used to have these issues, these are caused by aggresive coreclock/memclock/too low vddc voltage. Either/or combination of those 3 settings.

I solved the bsod by using more conservative core clock/mem clock and increase the vddc voltage

[Pennywis3]

Very bad expirience i have today.
Miner stopped mining with messege can't connect to the dev pool, you are blocked
Hey its not my problem if your dev pools not working, maybe you review your policy

You can generate a PS script to restart a process if its not active.
If you don't know PS you can always use a program to do it for you, you can check some here:
https://www.raymond.cc/blog/keep-application-running-by-automatically-rerun-when-closed/

Doctor, the miner sometimes closes anyway, even if its not blocked by FW or antivirus, you should look into this.

[doktor83]

Very bad expirience i have today.
Miner stopped mining with messege can't connect to the dev pool, you are blocked
Hey its not my problem if your dev pools not working, maybe you review your policy

You can generate a PS script to restart a process if its not active.
If you don't know PS you can always use a program to do it for you, you can check some here:
https://www.raymond.cc/blog/keep-application-running-by-automatically-rerun-when-closed/

Doctor, the miner sometimes closes anyway, even if its not blocked by FW or antivirus, you should look into this.

I would if i get some logs