Topic: SRBMiner-MULTI GPU & CPU Miner 0.9.4 - page 50. (Read 44750 times)
Is it possible to mine tellor now? Coz uupool no more support tellor coin.
Thank you for answers. How about to add cuckaroo/cuckatoo algos for cpu (grin/epic have some ocl and cpu plugins) and gpu?
Hi, not atm, but maybe once in the future. There are a lot of other algorithms i would like to add too
V0.4.7
+ Removed devfee for 'blake2s', 'mtp', 'keccak', 'yespoweritc', 'm7mv2', 'cryptonight_catalans', 'cryptonight_talleo' algorithms
+ Lowered 'bl2bsha3' devfee to standard fee (~0.85%)
+ Fixed 'randomx' algorithm when running on 64+ threads
+ Fixed config parameter "cpu_affinity"
- Now you get even more 0% devfee algorithms
- Thanks to @Zorg33 (https://github.com/doktor83/SRBMiner-Multi/issues/10) managed to fix 'randomx' too low hashrate when running with 64+ threads
- "cpu_affinity" parameter in configuration file was not working, now it is fixed
+ Removed devfee for 'blake2s', 'mtp', 'keccak', 'yespoweritc', 'm7mv2', 'cryptonight_catalans', 'cryptonight_talleo' algorithms
+ Lowered 'bl2bsha3' devfee to standard fee (~0.85%)
+ Fixed 'randomx' algorithm when running on 64+ threads
+ Fixed config parameter "cpu_affinity"
- Now you get even more 0% devfee algorithms
- Thanks to @Zorg33 (https://github.com/doktor83/SRBMiner-Multi/issues/10) managed to fix 'randomx' too low hashrate when running with 64+ threads
- "cpu_affinity" parameter in configuration file was not working, now it is fixed
Thank you for answers. How about to add cuckaroo/cuckatoo algos for cpu (grin/epic have some ocl and cpu plugins) and gpu?
Hi, Doktor. Any plans to add dual mining for RandomEpic/ProgPow for epic coin?
Another question - if i have one card with 8gb and other with 4gb can miner use first card 8gb memory
for DAG so second card with 4gb memory can continue mine eth after dag size >=4 gb?
Another question - if i have one card with 8gb and other with 4gb can miner use first card 8gb memory
for DAG so second card with 4gb memory can continue mine eth after dag size >=4 gb?
Hi, no dual mining is in plan at the moment.
And no, what you would like with the 8g-4g is not possible
Hi, Doktor. Any plans to add dual mining for RandomEpic/ProgPow for epic coin?
Another question - if i have one card with 8gb and other with 4gb can miner use first card 8gb memory
for DAG so second card with 4gb memory can continue mine eth after dag size >=4 gb?
Another question - if i have one card with 8gb and other with 4gb can miner use first card 8gb memory
for DAG so second card with 4gb memory can continue mine eth after dag size >=4 gb?
V0.4.6
+ Fixed 'defyx' algorithm on Intel CPU's (previous version broke it)
+ Hashrate stats (pressing 'h') now shows current hashing speed of CPU, and not the 10 second average
+ Lowered pool reconnection time to 15 sec
+ Added parameter '--gpu-ethash-mode' to cmd parameters
+ Added parameter 'gpu_ethash_mode' to config parameters
+ Added parameter 'ethash_mode' to gpu_conf parameters
+ Added parameter '--disable-cpu-auto-affinity'
- parameter ethash mode, if set to 1, will allow for a little more memory allocation of DAG on the GPU, so if you are really on the edge of the limit, this could help you mine with normal speed for few more epochs. The bad thing is, if your gpu is on pcie 1.0 (riser) then the DAG creation time will be very high, rx580 is about 50 minutes. Also you should only use it with Ethereum, because the devfee is mining ETH, so imagine if you mine ETC , it takes 50 minutes to created DAG, and then on devfee mining again 50 minutes to create DAG for ETH..
- parameter --disable-cpu-auto-affinity turns off the automatic affinity setting, and leaves that to the OS to manage.
+ Fixed 'defyx' algorithm on Intel CPU's (previous version broke it)
+ Hashrate stats (pressing 'h') now shows current hashing speed of CPU, and not the 10 second average
+ Lowered pool reconnection time to 15 sec
+ Added parameter '--gpu-ethash-mode' to cmd parameters
+ Added parameter 'gpu_ethash_mode' to config parameters
+ Added parameter 'ethash_mode' to gpu_conf parameters
+ Added parameter '--disable-cpu-auto-affinity'
- parameter ethash mode, if set to 1, will allow for a little more memory allocation of DAG on the GPU, so if you are really on the edge of the limit, this could help you mine with normal speed for few more epochs. The bad thing is, if your gpu is on pcie 1.0 (riser) then the DAG creation time will be very high, rx580 is about 50 minutes. Also you should only use it with Ethereum, because the devfee is mining ETH, so imagine if you mine ETC , it takes 50 minutes to created DAG, and then on devfee mining again 50 minutes to create DAG for ETH..
- parameter --disable-cpu-auto-affinity turns off the automatic affinity setting, and leaves that to the OS to manage.
Hi, Any plans to add KAWPOW (RVN) algo to your AMD miner?
Cheers
Cheers
Quote from: doktor83
what amd driver version for eth mining?
Has someone reported hashrate for a Ryzen 3600 (Random-X XMR) with this miner ? Just caught the thread, haven't read the whole content.
Hi Doktor, can we dual mine Epic and other RandomX coin while Epic RandomX is idle?
No, it's not working that way, sorry.
Hi Doktor, can we dual mine Epic and other RandomX coin while Epic RandomX is idle?
V0.4.5
+ Added 'randomepic' algorithm (EpicCash RandomX) solo/pool
+ 'ethash' and 'ubqhash' algorithm faster ~1.5% on Vega
+ 'RandomX' algorithm improvements
+ Fixed allocation of 'RandomX' dataset for 1Gb pages (--randomx-use-1gb-pages parameter)
+ Fixed an issue with MSR tweaks not being applied correctly
+ Lowered 'minotaur' algorithm devfee to 2%
+ Fixed crashing of 'minotaur' algorithm on AVX2 capable cpu's
+ '--send-stales' is now set per pool, it accepts true/false. Default is true.
+ 'send_stales' parameter added to pools config file too
+ '--gpu-intensity' option can now be used in combination with auto gpu setup in cmd
+ Improved gpu watchdog
+ Added temporary parameter '--cnpool-bbc-fix' which disables workers if job diff >= 50
+ Minor bug fixes
- When mining 'randomepic' miner workers will stop when jobs are not randomx.
- Found a bug that was here from the beginning of randomx addition, where MSR tweaks were applied incorrectly, so the hashrate was lower than it could be.
- '--send-stales' and send_stales parameters now accept a bool (true, false) and are now set per pool. Some pools accept stales while others don't, with this parameter you can control to which pool you want to send/not send shares that miner thinks are stale. Default is true, it means stale shares are sent if this parameter is not set to false.
- added parameter '--cnpool-bbc-fix' on request of cnpool.cc operator. If this parameter is used miner will stop workers when a job diff >= 50 received.
- Now you can use '--gpu-intensity' parameter alone, and miner will set the rest parameters for you automatically.
Example :
1. Set same intensity to every gpu:
2. Set different intensities (4 gpu's in rig):
Also from previous version the '--gpu-tweak-profile' parameter works the same, so it can be applied without setting the rest of the gpu parameters.
Example:
1. Apply intensity 24 to every gpu and set tweak-profile 5 to them
+ Added 'randomepic' algorithm (EpicCash RandomX) solo/pool
+ 'ethash' and 'ubqhash' algorithm faster ~1.5% on Vega
+ 'RandomX' algorithm improvements
+ Fixed allocation of 'RandomX' dataset for 1Gb pages (--randomx-use-1gb-pages parameter)
+ Fixed an issue with MSR tweaks not being applied correctly
+ Lowered 'minotaur' algorithm devfee to 2%
+ Fixed crashing of 'minotaur' algorithm on AVX2 capable cpu's
+ '--send-stales' is now set per pool, it accepts true/false. Default is true.
+ 'send_stales' parameter added to pools config file too
+ '--gpu-intensity' option can now be used in combination with auto gpu setup in cmd
+ Improved gpu watchdog
+ Added temporary parameter '--cnpool-bbc-fix' which disables workers if job diff >= 50
+ Minor bug fixes
- When mining 'randomepic' miner workers will stop when jobs are not randomx.
- Found a bug that was here from the beginning of randomx addition, where MSR tweaks were applied incorrectly, so the hashrate was lower than it could be.
- '--send-stales' and send_stales parameters now accept a bool (true, false) and are now set per pool. Some pools accept stales while others don't, with this parameter you can control to which pool you want to send/not send shares that miner thinks are stale. Default is true, it means stale shares are sent if this parameter is not set to false.
- added parameter '--cnpool-bbc-fix' on request of cnpool.cc operator. If this parameter is used miner will stop workers when a job diff >= 50 received.
- Now you can use '--gpu-intensity' parameter alone, and miner will set the rest parameters for you automatically.
Example :
1. Set same intensity to every gpu:
Code:
SRBMiner-MULTI.exe --algorithm ethash --gpu-intensity 24 --disable-cpu --pool eth-eu1.nanopool.org:9999 --wallet 0x74fb183Dfa951A9dC8C6de5ae4E1A8587fa6f371
2. Set different intensities (4 gpu's in rig):
Code:
SRBMiner-MULTI.exe --algorithm ethash --gpu-intensity 24,23,23,25 --disable-cpu --pool eth-eu1.nanopool.org:9999 --wallet 0x74fb183Dfa951A9dC8C6de5ae4E1A8587fa6f371
Also from previous version the '--gpu-tweak-profile' parameter works the same, so it can be applied without setting the rest of the gpu parameters.
Example:
1. Apply intensity 24 to every gpu and set tweak-profile 5 to them
Code:
SRBMiner-MULTI.exe --algorithm ethash --gpu-intensity 24 --gpu-tweak-profile 5 --disable-cpu --pool eth-eu1.nanopool.org:9999 --wallet 0x74fb183Dfa951A9dC8C6de5ae4E1A8587fa6f371
fixed. 56.2mh with intensity 25 according to miner stats.
We know that PM shows hashrate by ~2% higher than it really is, so you are now hashing approx with same speed as with PM.
-- gpu-intensity 23 (and others) does not apply (Windows, console). And if I have parallel mining on the processor (by another miner), the hashrate falls. However, xmrig has the lowest priority in the configuration.
If you want to manually set the gpu's you must use all mandatory parameters (--gpu-id, --gpu-intensity or --gpu-raw-intensity, --gpu-worksize and --gpu-threads)
So:
Code:
--gpu-id 0 --gpu-intensity 23 --gpu-worksize 64 --gpu-threads 1
Next release won't have this requirement
What's typical hash for rx5700 in this miner ?
It depends how you modded your card, but 5700 xt can do around 59mhs
just tried SRB :55.7 PM:57.3 in miner screen, 5700 non XT 1400 core 930 mem
Did you use default settings ? Because defaults are set too low (intensity 22), 24-25 should be a good value.
Code:
{
"algorithm" : "ethash",
"disable_cpu" : true,
"disable_gpu" : false,
"intensity" : 24,
}It's not "intensity" anymore, now it is "gpu_intensity"
What's typical hash for rx5700 in this miner ?
It depends how you modded your card, but 5700 xt can do around 59mhs
just tried SRB :55.7 PM:57.3 in miner screen, 5700 non XT 1400 core 930 mem
Did you use default settings ? Because defaults are set too low (intensity 22), 24-25 should be a good value.
Code:
{
"algorithm" : "ethash",
"disable_cpu" : true,
"disable_gpu" : false,
"intensity" : 24,
}It's not "intensity" anymore, now it is "gpu_intensity"
What's typical hash for rx5700 in this miner ?
It depends how you modded your card, but 5700 xt can do around 59mhs
just tried SRB :55.7 PM:57.3 in miner screen, 5700 non XT 1400 core 930 mem
Did you use default settings ? Because defaults are set too low (intensity 22), 24-25 should be a good value.
Code:
{
"algorithm" : "ethash",
"disable_cpu" : true,
"disable_gpu" : false,
"intensity" : 24,
}
-- gpu-intensity 23 (and others) does not apply (Windows, console). And if I have parallel mining on the processor (by another miner), the hashrate falls. However, xmrig has the lowest priority in the configuration.
What's typical hash for rx5700 in this miner ?
It depends how you modded your card, but 5700 xt can do around 59mhs
just tried SRB :55.7 PM:57.3 in miner screen, 5700 non XT 1400 core 930 mem
Did you use default settings ? Because defaults are set too low (intensity 22), 24-25 should be a good value.
Jump to: