Topic: SRBMiner-MULTI GPU & CPU Miner 0.9.4 - page 17. (Read 42480 times)

Is there a possibility to add extra threads for dual mining algorithms?

I noticed it was limited to 1 thread per algo when dual-mining.

For example I want 2threads for ETH/ETC and then one thread for HEAVYHASH.

--gpu-intensity 20+20;10 for example ignored the second thread for ETC/ETH. Only used the first value per algo.

V1.0.4
+ Improved performance on all dual mineable algorithms for RDNA2 GPU's
+ Minor bug fixes

Same as for every other algorithm, stratum only.

V1.0.3
+ Added algorithm 'mike' (Vkax coin) for CPU mining, fee 0.85%
+ Significantly improved performance on 'curvehash' algorithm for GPU's
+ Reduced devfee for 'kaspa' and 'heavyhash' to 0.85%
+ Using any of the supported dual mining combinations, automatically enables auto tune
+ Added possibility to disable an algorithm in dual mining mode per GPU*
+ Default value for '--gpu-dual-max-loss' is lowered from 5% to 3%*

***
It is now possible to disable one of the algorithms in dual mining mode per GPU.

--gpu-id ETHASH_GPUS;KASPA_GPUS [separator between GPU's is !, as usual]

For example you have a 4 GPU rig and want to dual mine ETH+KAS on GPU's 0,1,3 and ETH only on GPU 2 (without KAS) :

--gpu-id 0!1!2!3;0!1!3


***
--gpu-dual-max-loss parameter should be used to set the ratio between the hashrates of the two algorithms in dual mining mode.
Lower value gives higher priority to the first algorithm (ETH/ETC/AUTOLYKOS2).
As always, GPU CORE clock affects the effectiveness of this parameter. Sometimes a slightly higher CORE clock can make a big difference.

What about KASPA solo mining ?
... solo+tcp://127.0.0.1 doesn't work on my node ..

Devfee is mined on a different connection, not the same you are connected to the pool. Also you won't see any devfee shares sent on the screen/log - it's not displayed to the user.
What happened in your case that i don't know, but i am almost certain it's wasn't mining to my devfee wallet. There are also a few mechanisms that abort devfee mining if it might get stuck.

i think there is some bug in stbminer, that it forget to switch back to user pool and will be mining to developers wallet until miner is restarted. I have seen this problem a few times, 1-2 times every 1-2 months. Right now i have 21 cpus mining 2 different coins. Rigs are not local, some of them controlled with vnc. Two days ago one cpu worker dissappeared from pool, but i see that this rig is alive and gpus are online and working fine, just cpu worker is missing. Just got here and i see same problem as before - srbminer is running fine and shares sent all the time, but not to my wallet. Closed srbminer and run again - i see my worker on zergpool again as before. Version is 0.9.8, but i had same problem with previous versions.

Use lolminer if you want to mine KAS for example. But in my opinion it is not worth, if KAS not rise in price next time, i think it will die slowly. Profits so low

let's we give him one fury for tests...

 
Read the instructions, it is always added:

V1.0.0

+ Added ETHASH/ETCHASH + BLAKE3_ALEPHIUM dual mining mode
+ Added AUTOLYKOS2 + BLAKE3_ALEPHIUM dual mining mode
+ Fixed ETHASH/ETCHASH + KASPA mining on nicehash
+ Added parameter '--gpu-table-slow-build' which slows down DAG/Datatable creation for ETHASH/ETCHASH/UBQHASH/PROGPOW/AUTOLYKOS2 algorithms (might help in avoiding crash)
+ Parameter '--gpu-dual-mode' isn't needed anymore, miner will auto check if selected algorithms are eligible for dual mining
+ Fixed a few bugs related to dual mining mode

Dual mining tips:

Same as before, it is important to follow this simple rule:

If you increase CORE clock of the GPU because you want higher hashrate on A1 (KASPA/ALEPHIUM), you probably should also increase intensity of the SECOND algorithm too!
Not doing so might leave you on the same hashrate even if you increased the CORE clock!

Example:

Leave intensity of A0 on auto, and set intensity of A1 to 4 : --gpu-intensity 0 --gpu-intensity 4 , or --gpu-intensity 0;4

Dual ethereum/alephium example:

--disable-cpu --algorithm ethash --algorithm blake3_alephium --pool ethereum-pool-here --pool alephium-pool-here --wallet ethereum-wallet-here --wallet alephium-wallet-here

Dual ergo/alephium example:

--disable-cpu --algorithm autolykos2 --algorithm kaspa --pool ergo-pool-here --pool alephium-pool-here --wallet ergo-wallet-here --wallet alephium-wallet-here


Support for old gpu´s was dropped in version 0.9.5 (read the changes) because doktor has nothing for tests.

what a f..k man? Where are support for fiji? in many algos they still strong, one example its 23mh on kawpow and many other. many years we works with your miner

Hi.
Do you have plans to add 'ethash + blake3b' ?
Membercash, not a bad coin on blake3b https://bitcointalksearch.org/topic/--5399940

V1.0.2

+ Added ETHASH/ETCHASH + HEAVYHASH (OBTC/PBTC) dual mining mode
+ Added AUTOLYKOS2 + HEAVYHASH dual mining mode
+ Faster DAG creation on RDNA2 GPU's
+ Possible fix for case when some RDNA/RDNA2 GPU's won't create DAG for ETHASH/PROGPOW algorithms on Linux based OS

Sorry to bother you, but can you create such version to mining in solo like in good old days? Or it is too hard to realize technically ?

No

Hey! Is it possible to use SRBMiner for solo mining without stratum?

V1.0.1

+ Added auto tune for dual mining (use --gpu-auto-tune 1 to enable)*
+ Added parameter '--gpu-dual-max-loss' which defines the maximum loss of A0 hashrate we allow while auto tuning in dual mode, defined in percent
+ Minor bug fixes

***
Auto tune for dual mode:

Auto tune will try to find the best intensity value for the SECOND algorithm, which heavily depends on the CORE clock of your GPU.
Higher the clock, probably higher the intensity value will be.

You can set the maximum loss of hashrate (in percent) you allow for the FIRST algorithm by using the '--gpu-dual-max-loss' parameter which accepts a number between 1 and 90.
Default for this value, if not set, is 5%.

Dual ethereum/kaspa example [allow max 10% loss in hashrate for ethereum]:

V1.0.0

+ Added ETHASH/ETCHASH + BLAKE3_ALEPHIUM dual mining mode
+ Added AUTOLYKOS2 + BLAKE3_ALEPHIUM dual mining mode
+ Fixed ETHASH/ETCHASH + KASPA mining on nicehash
+ Added parameter '--gpu-table-slow-build' which slows down DAG/Datatable creation for ETHASH/ETCHASH/UBQHASH/PROGPOW/AUTOLYKOS2 algorithms (might help in avoiding crash)
+ Parameter '--gpu-dual-mode' isn't needed anymore, miner will auto check if selected algorithms are eligible for dual mining
+ Fixed a few bugs related to dual mining mode

Dual mining tips:

Same as before, it is important to follow this simple rule:

If you increase CORE clock of the GPU because you want higher hashrate on A1 (KASPA/ALEPHIUM), you probably should also increase intensity of the SECOND algorithm too!
Not doing so might leave you on the same hashrate even if you increased the CORE clock!

Example:

Leave intensity of A0 on auto, and set intensity of A1 to 4 : --gpu-intensity 0 --gpu-intensity 4 , or --gpu-intensity 0;4

Dual ethereum/alephium example:

--disable-cpu --algorithm ethash --algorithm blake3_alephium --pool ethereum-pool-here --pool alephium-pool-here --wallet ethereum-wallet-here --wallet alephium-wallet-here

Dual ergo/alephium example:

--disable-cpu --algorithm autolykos2 --algorithm kaspa --pool ergo-pool-here --pool alephium-pool-here --wallet ergo-wallet-here --wallet alephium-wallet-here

0.9.9

+ Improved performance of KASPA mining in dual mode for ETH/ETC+KASPA and AUTOLYKOS2+KASPA for some GPU's [ read instructions below* ]
+ Added algorithm 'curvehash' for GPU mining*
+ Fixed compatibility of KawPow with miningpoolhub pool


***
Dual mining tuning:

Intensity of the second algorithm (KASPA) affects how much time should be spent on mining the second algorithm (KASPA).
Value of intensity goes from 1-31 , but everything over 8 will probably only degrade performance of both algorithms too much.

There is only one tip : if you want more KASPA you will need to increase CORE clock AND also the intensity of the second algorithm (KASPA)
Increasing only CORE clock without setting a higher intensity value probably won't increase KASPA hashrate.

Some GPU's can use only predefined intensities:

Polaris : 1, 2, 4
Rx550/Rx560 : 1

***
Curvehash GPU:

This is a quick release - it was tested on Windows on drivers < 22.x. and it should work ok.
Linux and mining os-es were barely tested - polaris works - not sure about the rest.. Rocm is not supported.

If a pool has a very low start diff. you should use parameter '--diff-factor' so you shouldn't flood the pool with shares.

OC Tip: MEM clock should go down as much as possible, CORE should go up as much as possible

V0.9.8

+ Added ETHASH/ETCHASH + KASPA dual mining mode
+ Added AUTOLYKOS2 + KASPA dual mining mode
+ Performance increase on 'curvehash' algorithm for CPU ( up to ~40% )
+ Added parameter '--gpu-dual-mode' which must be used to enable the new dual mining mode
+ Upgraded command line parser*
+ Bug fixes

***
To maximise your mining profits, SRBMiner added real dual mining of :

ETHASH/ETCHASH + KASPA
AUTOLYKOS2 + KASPA

This feature is still in experimental/beta phase, so feedback is welcome.

Setup procedure for the new dual mining mode is exactly the same as before, you just need to append parameter '--gpu-dual-mode' to enable the new mode.

You can manually set the intensities for both algorithms if you wish, it can have a positive/negative impact on the hashrates.
An auto tuning mode for the new dual mining is planned to be added in the future.
Currently using the '--gpu-auto-tune' parameter won't do anything.

Dual ethereum/kaspa example:

--disable-cpu --algorithm ethash;kaspa --gpu-dual-mode --pool ethereum-pool-here;kaspa-pool-here --wallet ethereum-wallet-here;kaspa-wallet-here

Dual ergo/kaspa example:

--disable-cpu --algorithm autolykos2;kaspa --gpu-dual-mode --pool ergo-pool-here;kaspa-pool-here --wallet ergo-wallet-here;kaspa-wallet-here


***
Upgraded command line parser:

Order of parameters doesn't matter anymore, parser will internally fix ordering.
Also there is a new way of defining multiple algorithms, parameters will be merged.

Example :
--algorithm ethash;kaspa can be written as --algorithm ethash --algorithm kaspa

Rule is that first defined parameter will be defined for algorithm1, same parameter if used will be defined for algorithm2 , etc..

Example:
--algorithm ethash;kaspa --gpu-intensity 24;23 can be written as --algorithm ethash --gpu-intensity 24 --algorithm kaspa --gpu-intensity 23


Example for previous ethereum/kaspa can be written as (ex.):

--gpu-dual-mode --algorithm ethash --pool ethereum-pool-here --wallet ethereum-wallet-here --algorithm kaspa --pool kaspa-pool-here --wallet kaspa-wallet-here --disable-cpu