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

[lebuawu2]

why so angry?

i read this topic last 2 weeks

intensity 60 = 800hs, kill process and start with 56 = 1200hs

thanks

what is your cclock and mclock? is it blockchain driver or 18.3.4?

[nordmann666]

why so angry?

i read this topic last 2 weeks

intensity 60 = 800hs, kill process and start with 56 = 1200hs

thanks

what is your cclock and mclock? is it blockchain driver or 18.3.4?

1408/1099 water cooled so hbm around 50°C

[lebuawu2]

why so angry?

i read this topic last 2 weeks

intensity 60 = 800hs, kill process and start with 56 = 1200hs

thanks

what is your cclock and mclock? is it blockchain driver or 18.3.4?

1408/1099 water cooled so hbm around 50°C

with that setting should able to get 1400+ or 1500+, maybe you need to post your config also.

[nordmann666]

"gpu_conf" :
[
   { "id" : 1, "intensity" : 56, "worksize" : 16, "threads" : 2}
]
}

tried with different worksize


-----------------------------------------------------

set GPU_MAX_ALLOC_PERCENT 100
set GPU_USE_SYNC_OBJECTS 1
set GPU_MAX_HEAP_SIZE 100
set GPU_SINGLE_ALLOC_PERCENT 100
set GPU_FORCE_64BIT_PTR 1

@echo off

cls
SETLOCAL EnableExtensions
set EXE=SRBMiner-CN_svega.exe

:START
FOR /F %%x IN ('tasklist /NH /FI "IMAGENAME eq %EXE%"') DO IF %%x == %EXE% goto RUNNING
%EXE% --config config_s2.txt --pools pools_h.txt
TIMEOUT /T 15 /NOBREAK
goto START
:RUNNING
echo "%EXE% is already running"
pause

-----------------------------------------------------

why so angry?

i read this topic last 2 weeks

intensity 60 = 800hs, kill process and start with 56 = 1200hs

thanks

what is your cclock and mclock? is it blockchain driver or 18.3.4?

1408/1099 water cooled so hbm around 50°C

with that setting should able to get 1400+ or 1500+, maybe you need to post your config also.

[wgestickley]

Doc, I can confirm, that my RX580 cards with 8GB get a Hashrate drop after Pool switch / dev Pool. After start and close GPU-Z while miner is running, the max. Hashrate come back again.

Would be nice to implement the „Magic Function“ to the Miner :-)

Greetings from Germany
Oliver

[istr]

Hello guys,

In heavy algorithm I use 1300 mhz core and depending on memory type (Samsung, Micron...) some cards work on 2000 mhz and some others on 2150 - 2200.
What do you suggest for voltages on both core and memory?
Right now I use core 975 mv and mem 1000 mv. Is it good or I can lower it a little more to save energy?
Cards are RX 580 8gb.

Thanks in advance!

[doktor83]

Doc, I can confirm, that my RX580 cards with 8GB get a Hashrate drop after Pool switch / dev Pool. After start and close GPU-Z while miner is running, the max. Hashrate come back again.

Would be nice to implement the „Magic Function“ to the Miner :-)

Greetings from Germany
Oliver

Hashrate shouldn't drop after devfee mining, because when switching do devfee pool, user pool is not disconnected, it's running in the background, and when devfee finishes it just takes work from user pool and continues hashing. User pool can lower difficulty because user is not sending shares for 1 minute, but i don't believe that should be a reason for the hashrate to drop.

Also it would be nice to know what is gpuz doing

[wgestickley]

Doc, I can confirm, that my RX580 cards with 8GB get a Hashrate drop after Pool switch / dev Pool. After start and close GPU-Z while miner is running, the max. Hashrate come back again.

Would be nice to implement the „Magic Function“ to the Miner :-)

Greetings from Germany
Oliver

Hashrate shouldn't drop after devfee mining, because when switching do devfee pool, user pool is not disconnected, it's running in the background, and when devfee finishes it just takes work from user pool and continues hashing. User pool can lower difficulty because user is not sending shares for 1 minute, but i don't believe that should be a reason for the hashrate to drop.

Also it would be nice to know what is gpuz doing

Tricky :-)

I also think, that the lower difficulty after devfee is NOT the reason because i set a static Diff.

Algo: Heavy, last Version of GPU-Z

[hesido]

Doc, I can confirm, that my RX580 cards with 8GB get a Hashrate drop after Pool switch / dev Pool. After start and close GPU-Z while miner is running, the max. Hashrate come back again.

Would be nice to implement the „Magic Function“ to the Miner :-)

Greetings from Germany
Oliver

Hashrate shouldn't drop after devfee mining, because when switching do devfee pool, user pool is not disconnected, it's running in the background, and when devfee finishes it just takes work from user pool and continues hashing. User pool can lower difficulty because user is not sending shares for 1 minute, but i don't believe that should be a reason for the hashrate to drop.

Also it would be nice to know what is gpuz doing

I will go through logs sometime but devfee switch does not seem to cause trouble with the hashrate, and on my end, keeping GPU-Z turned on keeps everything at full speed. Having said that, manual switch shoots down the hashrate a lot. Just having GPU-Z open fixes that too in a few minutes.

I wonder why you cannot see the effect, maybe you are running a special debug version doing extra calls or something?

Hello guys,

In heavy algorithm I use 1300 mhz core and depending on memory type (Samsung, Micron...) some cards work on 2000 mhz and some others on 2150 - 2200.
What do you suggest for voltages on both core and memory?
Right now I use core 975 mv and mem 1000 mv. Is it good or I can lower it a little more to save energy?
Cards are RX 580 8gb.

Thanks in advance!

Istr, what is your hashrate per rx580 card?

I do 1160mhz coreclock + 2220mhz on my microns, 880mV vcore and vmem., I get 1010 mh/s, using polaris 1 click timings
I do 1160mhz coreclock + 2100mhz on one of my hynix, 900 mV vcore and vmem, with a custom timing strap, and get 1005mh/s,
I do 1160mhz coreclock + 2190mhz on my other hynix, 900 mV vcore and vmem, with polaris 1 click timings, and get 950mh/s

I think, you cannot set vcore to be less than vmem, so your cores are effectively running at 1000mV.

I must say though, my hynix that does 1005 is not very stable, it does crash every once in a while, and it is the one with hashing errors, although no memory errors are reported. Some of my microns report mem errors but they never procuded bad hashes. That hynix one was bad from the beginning, but the custom ram timings saved a lot of performance and reduced the errors to acceptable levels.

I can't flash the other hynix as it's index is too high and I have to shutdown mining and disable some gpu's to reach it with ATI flash.

[argai666]

Is it possible to start SRBMiner while not using config.txt and pools.txt like for example sgminer? Something like: "SRBMiner -pool <pool> -user <wallet> -i <intensity> -threads <number of threads> and so on. Can this be done?

like you can do in claymore? is that what you're talking about?

Exactly! This will make switching pools much easier.

Did you really use Claymore? Last time i checked, you need to use different pools.txt file so you can use pool switch option. So what is the difference?

[hesido]

Is it possible to start SRBMiner while not using config.txt and pools.txt like for example sgminer? Something like: "SRBMiner -pool <pool> -user <wallet> -i <intensity> -threads <number of threads> and so on. Can this be done?

like you can do in claymore? is that what you're talking about?

Exactly! This will make switching pools much easier.

Did you really use Claymore? Last time i checked, you need to use different pools.txt file so you can use pool switch option. So what is the difference?

I do not understand what the big deal is either, create a pools-a.txt, pools-b.txt, and have separate .bat's targeting those pool files.

Having said that, it would be nice to have a dynamic reload of the selected pools txt.

[rusman2]

Here the best results are at 60 intensity and just one thread. At 1230/2020 I can reach 775~785H/s in my RX570 4GB 4GB Elpida.

How are you able to run 60 intensity, 8 worksize on a 4GB card?
I am only able to run at 51 intensity with 8 worksize. At 52 I get a scratch pad size error, meaning not enough physical mem on the card.
I'm getting around 700H/s for two Hynix cards at 1350/2000.

[bulgar73]

Here the best results are at 60 intensity and just one thread. At 1230/2020 I can reach 775~785H/s in my RX570 4GB 4GB Elpida.

How are you able to run 60 intensity, 8 worksize on a 4GB card?
I am only able to run at 51 intensity with 8 worksize. At 52 I get a scratch pad size error, meaning not enough physical mem on the card.
I'm getting around 700H/s for two Hynix cards at 1350/2000.

same for me with MSI RX570 4GB, max.51 intensity, 52 is error ...

[heavyarms1912]

Cryptonighy-Heavy.  1200+ H/s on RX 570 8gb (Micron).  Also, serving as primary display card so couldn't push them too far.  Makes my Vega 56 look bad hashing at 1400 H/s 



have left them hashing at 1130 H/s 1250/2200 clocks @ 906mV.

[Q_R_V]

I do not understand what the big deal is either, create a pools-a.txt, pools-b.txt, and have separate .bat's targeting those pool files.

[...]

Let's say i want to remotely switch pool to new one (not defined in in pools) and algo to different while using only one single command (well actually two, because i would need to kill mining process first). Currently  pools.txt and config.txt needs to be recreated on the target rig. Next miner process is killed and restarted, becouse in my case it seems to be easier.
I'm not calling to abandon pools.txt and config.txt files, i'm just gauging interest around new feature.
Ability to start new process with a parameter seems cleaner and more logical to me.
Is it really that perverted in your eyes?

[goodminer]

Hmm something weird is happening when I tried to mine iridiumcoin for example litev7 algo, it gives shares non stop in burst, but statistic on pool tells that my r7 370 on def settings gives me about 200 h/s only? What is wrong, is this algo bad optimized, or what?
I tried last version of serbminer.

[lebuawu2]

Cryptonighy-Heavy.  1200+ H/s on RX 570 8gb (Micron).  Also, serving as primary display card so couldn't push them too far.  Makes my Vega 56 look bad hashing at 1400 H/s 



have left them hashing at 1130 H/s 1250/2200 clocks @ 906mV.

which intensity you used? did you modded your bios with polaris editor one click timing?

[istr]

Doc, I can confirm, that my RX580 cards with 8GB get a Hashrate drop after Pool switch / dev Pool. After start and close GPU-Z while miner is running, the max. Hashrate come back again.

Would be nice to implement the „Magic Function“ to the Miner :-)

Greetings from Germany
Oliver

Hashrate shouldn't drop after devfee mining, because when switching do devfee pool, user pool is not disconnected, it's running in the background, and when devfee finishes it just takes work from user pool and continues hashing. User pool can lower difficulty because user is not sending shares for 1 minute, but i don't believe that should be a reason for the hashrate to drop.

Also it would be nice to know what is gpuz doing

Tricky :-)

I also think, that the lower difficulty after devfee is NOT the reason because i set a static Diff.

Algo: Heavy, last Version of GPU-Z

I can confirm what Doktor says, that there is NO drop in hashrate after devfee pool disconnects, some versions now.
That was the occasion some versions before when Doc was disconnecting user pool being connected to his.
But we have hashrate drop whenever we switch pools. That's a different story.
As you understand the problem arises whenever GPUs loose "load" and have to start over again.
For me, when my GPUs reach maximum speed with whatever way (GPU-Z,...), they stay ROCK STABLE for days
until I'll stop SRBminer  just to reboot my machine.
I don't switch pools now for long time. I'm rock stable in one :-)))

"Also it would be nice to know what is gpuz doing "
Doc, maybe if you ask the dev team?

[heavyarms1912]

Cryptonighy-Heavy.  1200+ H/s on RX 570 8gb (Micron).  Also, serving as primary display card so couldn't push them too far.  Makes my Vega 56 look bad hashing at 1400 H/s  



have left them hashing at 1130 H/s 1250/2200 clocks @ 906mV.

which intensity you used? did you modded your bios with polaris editor one click timing?

56. Dual threads.  Not the PBE strap.  PBE timings are pretty good too. They do 1080 on similar settings.  Haven't tested those at 1400+ clocks

[istr]

Hello guys,

In heavy algorithm I use 1300 mhz core and depending on memory type (Samsung, Micron...) some cards work on 2000 mhz and some others on 2150 - 2200.
What do you suggest for voltages on both core and memory?
Right now I use core 975 mv and mem 1000 mv. Is it good or I can lower it a little more to save energy?
Cards are RX 580 8gb.

Thanks in advance!

Istr, what is your hashrate per rx580 card?

I do 1160mhz coreclock + 2220mhz on my microns, 880mV vcore and vmem., I get 1010 mh/s, using polaris 1 click timings
I do 1160mhz coreclock + 2100mhz on one of my hynix, 900 mV vcore and vmem, with a custom timing strap, and get 1005mh/s,
I do 1160mhz coreclock + 2190mhz on my other hynix, 900 mV vcore and vmem, with polaris 1 click timings, and get 950mh/s

I think, you cannot set vcore to be less than vmem, so your cores are effectively running at 1000mV.

I must say though, my hynix that does 1005 is not very stable, it does crash every once in a while, and it is the one with hashing errors, although no memory errors are reported. Some of my microns report mem errors but they never procuded bad hashes. That hynix one was bad from the beginning, but the custom ram timings saved a lot of performance and reduced the errors to acceptable levels.

I can't flash the other hynix as it's index is too high and I have to shutdown mining and disable some gpu's to reach it with ATI flash.

[/quote]




Hi hesido,

thank you for your replay!

As I said in an earlier post I get from 1070 to 1148 h/s.
I have 2 rigs.

The first one with 6 RX. 2 480's and 4 580's 8gb all. (All 8gb in one rig for testing reasons).

2 Pulse's with Micron running 1300 mhz core and 2150 and 2200 mhz mem. I get on them 1070 and 1094 h/s stable.
1 XFX RX 480 8gb Samsung running 1220 mhz and 2000 mhz mem. I get on this 1106 h/s stable.
1 PowerColor RX 580 Micron running 1300 mhz core and 2170 mhz mem. I get on this 1080 h/s stable.
1 RX 480 Nitro+ Samsung running 1280 core and 2000 mem giving me 1134 h/s also stable.
and 1 Pulse RX 580 Samsung running 1300 core and 2000 mem giving me 1148 h/s also very stable for days and days.

Second rig and last, pitty!

4 Sapphire RX 570 4gb Nitro+ Hynix mem type running 1280 core and 2000 mem giving me 755 to 765 h/s also stable all the time.
and 2 Sapphire RX 580 4gb Nitro+ Elpida mem type running 1280 core and 1900 mem giving me 776 to 780 h/s stable all the time.
This second rig with the 4gb memory NEVER had the hashrate drop problem.
They reach maximum speed from the beginning.

I wish I had more!!! :-)))
To be honest I'm waiting for one Sapphire RX 580 8gb Special Edition. We'll see how it'll go.
All these on heavy algo.

Greetings from Hellas!