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

[daotienao1217]

download 1.4.3 if you really had devfee problem its fixed.
Also if devfee got stuck you didn't have those :

[2018-04-17 10:52:16] Pool sent a new job (ID: 606098711433573)

like hitmandn had.

[2018-04-17 10:53:05] Pool difficulty: 1531728

 this is an insanely high diff the pool sets, it's a badly setup pool.
Try to set a static diff on the pool if they have that option.

With version 1.4.3 I still get devfree errors when mining ipbc with static diff

[hitmandn]

in start.bat find this line :

%EXE% --config config.txt --pools pools.txt

and make it look like this :

%EXE% --config config.txt --pools pools.txt --logfile log.txt

then if this devfee thing happens again, i mean if it won't connect back to user pool, leave it for a few minutes and then send me the log.txt

My log file after 2 hours

http://www.mediafire.com/file/x19a9nxant39qfh/log.txt

[FFI2013]

I have a mixed rig with older gpus r9 270,280,390 how is the performance for these cards or if anyone is running them can you share your rates thank you

[dohfish]

Speed seems good, but the miner is simply nowhere near stable.

Crashes 8-10 times per hour, on basically all of my rigs - I cannot press "h" to show hashrate, that instant crashes the application.

All my rigs are perfectly stable on basically all other miners.

GGS, CastXMR, XMR-Stak, Claymore but not on this.

In terms of speed, judging from looking at "stats" which works, it is on par with GGS for performance, so im sticking with that, will test again later if things are more stable.

[cryptosize]

Hello everyone!

Does this miner support TeraScale/VLIW Radeon GPUs?

it should work, but you wont get speeds like on old claymore.
Which card do you have?

Radeon 5770 1GB GDDR5.

And yeah, hashing speed is only half compared to Claymore v9.7...

download v 1.4.3 and add --usekernel 3 in the bat like this :

%EXE% --config config.txt --pools pools.txt --usekernel 3

this is a test kernel im working on for those oldies

I guess this emulates the Claymore v9.7 kernel?

Thanks for caring about old GPUs (most devs don't), I'll give it a try!

I just tried it and I still only get 80 H/s compared to 176 H/s on Claymore v9.7... intensity is 14 (higher doesn't work), worksize 8, 1 thread.

Another problem with all miners I've tried (except Claymore v9.7) is that there is no "low intensity" option, which means the UI is lagging a lot (if you want to use your PC while mining).

I wish someone could replicate Claymore v9.7 characteristics and I'd gladly pay a mining fee...

[alczaty]

ipbc
rx580 8gb 2100h/s 1250/900
rx570 4gb 1850h/s 1100/900

GJ !

[heavyarms1912]

@doktor83 another feature request.  Can we set a custom donation level and custom dev mining time.  

eg. Like user config from
every ~2 hours 1 minute mining for the dev
can be changed to
every ~4 hours 3 minutes mining.

[JuanHungLo]

There may be some tweaking needed but I am very pleased with my Rx470's 8GB (Samsung) at 1260/2000 averaging 1014 H/s with an UberMix 3.1 strap with Monerov7.  All 40 cards are stable, especially after manually setting difficulty as the payment ID extension.  Really looking forward to API so I can use my Awesome Miner software with it.

[Larvitar]

There may be some tweaking needed but I am very pleased with my Rx470's 8GB (Samsung) at 1260/2000 averaging 1014 H/s with an UberMix 3.1 strap with Monerov7.  All 40 cards are stable, especially after manually setting difficulty as the payment ID extension.  Really looking forward to API so I can use my Awesome Miner software with it.

What intensity/threads are you using?

Ajusting manually can improve it, but no much. Mine 580 8GB Samsung can do around 1050~1060H/s in V7 (is better in Heavy) at 1280/2010.

[JuanHungLo]

There may be some tweaking needed but I am very pleased with my Rx470's 8GB (Samsung) at 1260/2000 averaging 1014 H/s with an UberMix 3.1 strap with Monerov7.  All 40 cards are stable, especially after manually setting difficulty as the payment ID extension.  Really looking forward to API so I can use my Awesome Miner software with it.

What intensity/threads are you using?

Ajusting manually can improve it, but no much. Mine 580 8GB Samsung can do around 1050~1060H/s in V7 (is better in Heavy) at 1280/2010.

I set it at 0, the software is choosing 57. 2 threads.
What I like to do is run them for about a week, check for bugs, then start moving them up in intensity, clocks, etc.
Problem is, the dev keeps updating the software!  That's a good problem, too.

[Larvitar]

Hi dorktor83. I'm stick at your miner, although have AwesomeMiner. Ya, SRBMiner is kicking asses!  

But patrike spoke if SRBMiner has a working API he could add it in AwesomeMiner support (would be great). Do you have plans to make an API for SRBMiner?

thanks
yes , i have it in plan, smos dev is also waiting on me to make an API

Thank you

I tested yesterday the 1.4.2 version (I didn't see the 1.4.3 before go to work). Ya, it's faster! The tunning is a bit different from 1.3.x, but with some intensity play everything is better than expected.
RX570 4GB at 1200/2020 996H/s at Cryptonight v7
RX580 8GB (Samsung) 1145H/s at Cryptonight Heavy

RX580 8gb 1145h/s on CN heavy? surely not.

 

Surely Yes
https://imgur.com/a/QgGhV

 

[mul]

AND FINALLY

i found what the devfee bug is.
Gonna fix it ASAP!

Hi doktor83! But the problem is not solved

[2018-04-17 21:03:42] DevFee stops in 10 seconds
[2018-04-17 21:03:43] DevFee stops in 9 seconds
[2018-04-17 21:03:44] DevFee stops in 8 seconds
[2018-04-17 21:03:45] DevFee stops in 7 seconds
[2018-04-17 21:03:46] DevFee stops in 6 seconds
[2018-04-17 21:03:47] DevFee stops in 5 seconds
[2018-04-17 21:03:48] DevFee stops in 4 seconds
[2018-04-17 21:03:49] DevFee stops in 3 seconds
[2018-04-17 21:03:50] DevFee stops in 2 seconds
[2018-04-17 21:03:52] DevFee stops in 7202 seconds
[2018-04-17 21:03:53] DevFee stops in 7201 seconds
[2018-04-17 21:03:54] DevFee stops in 7200 seconds
[2018-04-17 21:03:55] DevFee stops in 7199 seconds
[2018-04-17 21:03:56] DevFee stops in 7198 seconds
[2018-04-17 21:03:57] DevFee stops in 7197 seconds
[2018-04-17 21:03:58] DevFee stops in 7196 seconds
[2018-04-17 21:03:59] DevFee stops in 7195 seconds
[2018-04-17 21:04:00] DevFee stops in 7194 seconds
[2018-04-17 21:04:01] DevFee stops in 7193 seconds
[2018-04-17 21:04:02] DevFee stops in 7192 seconds
[2018-04-17 21:04:03] DevFee stops in 7191 seconds
[2018-04-17 21:04:04] DevFee stops in 7190 seconds
[2018-04-17 21:04:05] DevFee stops in 7189 seconds
[2018-04-17 21:04:06] DevFee stops in 7188 seconds...

[hiron1986]

download v 1.4.3 and add --usekernel 3 in the bat like this :

%EXE% --config config.txt --pools pools.txt --usekernel 3

this is a test kernel im working on for those oldies

What usekernel parameter do you recommend for RX 550?

Because when I press H, the hashrate is unstable. It gets to 375 h/s, then NaN h/s, then 180 h/s. Already tried with double thread and single. I already disable ADL (enabling it will crash the miner if I press H).

[coke15]

hi

wtf with 1.4.3?  0 shares arrive to pool!! but i have accepted.

i update my 1.2.4 and 1.3.2  to 1.4.3,i loose 40 min!
config txt is the same as 1.2.4 ans 1.3.2

what's going on?

come back to 1.3.2 shares are good on pool !

[doktor83]

hi

wtf with 1.4.3?  0 shares arrive to pool!! but i have accepted.

i update my 1.2.4 and 1.3.2  to 1.4.3,i loose 40 min!
config txt is the same as 1.2.4 ans 1.3.2

what going on?

read first page

[doktor83]

download v 1.4.3 and add --usekernel 3 in the bat like this :

%EXE% --config config.txt --pools pools.txt --usekernel 3

this is a test kernel im working on for those oldies

What usekernel parameter do you recommend for RX 550?

Because when I press H, the hashrate is unstable. It gets to 375 h/s, then NaN h/s, then 180 h/s. Already tried with double thread and single. I already disable ADL (enabling it will crash the miner if I press H).

you need to lower intensity, this kernel parameter thing wont help you

[coke15]

hi

wtf with 1.4.3?  0 shares arrive to pool!! but i have accepted.

i update my 1.2.4 and 1.3.2  to 1.4.3,i loose 40 min!
config txt is the same as 1.2.4 ans 1.3.2

what going on?

read first page

rhaaaaaaaaaaaaaaa :/

thanks to you.

edit:dev has 40 min 20k hashrate

[Larvitar]

ipbc
rx580 8gb 2100h/s 1250/900
rx570 4gb 1850h/s 1100/900

GJ !

Which are your settings?

Damn, the fourth different config.txt lol

EDIT: intensity 114, 2 threads
After some time it reaches 2150H/s. Wow

[Littlegrand]

Just checked my Miner's logs all look good except this one:

[2018-04-15 13:01:46] Switching back to user pool in 11 ticks
[2018-04-15 13:01:46] Switching back to user pool in 10 ticks
[2018-04-15 13:01:47] pool_have_job: Switching devfee process to new job
[2018-04-15 13:01:47] Switching back to user pool in 9 ticks
[2018-04-15 13:01:47] Switching back to user pool in 8 ticks
[2018-04-15 13:01:48] Switching back to user pool in 7 ticks
[2018-04-15 13:01:48] Switching back to user pool in 6 ticks
[2018-04-15 13:01:49] Switching back to user pool in 5 ticks
[2018-04-15 13:01:49] pool_have_job: Switching devfee process to new job
[2018-04-15 13:01:49] Switching back to user pool in 4 ticks
[2018-04-15 13:01:50] Switching back to user pool in 3 ticks
[2018-04-15 13:01:50] Switching back to user pool in 2 ticks
[2018-04-15 13:01:51] Switching back to user pool in 1 ticks
[2018-04-15 13:01:51] pool_have_job: Switching devfee process to new job
[2018-04-15 13:01:52] Switching back to user pool in 0 ticks
[2018-04-15 13:01:52] switch_pool: Disconnected from devfee pool
[2018-04-15 13:01:52] switch_pool: Connecting to user pool
[2018-04-15 13:01:52] Switching back to user pool in 14401 ticks
[2018-04-15 13:01:53] Switching back to user pool in 14400 ticks
[2018-04-15 13:01:53] Switching back to user pool in 14399 ticks


--J

[R0land]

download v 1.4.3 and add --usekernel 3 in the bat like this :

%EXE% --config config.txt --pools pools.txt --usekernel 3

this is a test kernel im working on for those oldies

What usekernel parameter do you recommend for RX 550?

Because when I press H, the hashrate is unstable. It gets to 375 h/s, then NaN h/s, then 180 h/s. Already tried with double thread and single. I already disable ADL (enabling it will crash the miner if I press H).

you need to lower intensity, this kernel parameter thing wont help you

Try Intensenty 25 or 26, Worksize 16, Threads 2.
My RX550 makes 464 h/s constant. @ 21W