Releases: doktor83/SRBMiner-Multi
SRBMiner-MULTI CPU & AMD GPU Miner 0.9.1 Win64 & Linux
9771d33a7043615e8f0c6eb4eb3c916b *SRBMiner-Multi-0-9-1-win64.zip
aea91d5c09d8ad7ad3f230f005ce537f *SRBMiner-Multi-0-9-1-Linux.tar.xz
- Added algorithm 'dynamo' (DYNAMO - Dynamo coin) for CPU mining, fee 3.00% (~20x-24x (+2000%) faster than the available public miner)
- Small performance increase on 'blake3_alephium' algorithm for CPU's with AVX2
- Added '--gpu-auto-tune 3' mode (auto tuning goes through every algorithm, but tuning one at a time - not all simultaneously)
- If creating DAG/Dataset but mining multiple algorithms, miner now pauses work on other algorithms while the DAG/Dataset is created
- Possible fix for 'not connected to pool' bug that appeared in previous version (0.9.0)
- Better auto setup values for eth/etc + alph 'dual' mining
- Added a workaround for 'blake3_alephium' algorithm pools that send both set_difficulty and set target in job
- Minor bug fixes
'--gpu-auto-tune 3' might help in finding good '--gpu-intensity' values when doing 'dual' mining.
Set your prioritized algorithm to algorithm 0 (ex. ETH + ALPH, not ALPH + ETH)
'dynamo' will have a lot of rejected shares with the message 'duplicate share'. That is because everyone is working on the same job. It's not a miner issue.
SRBMiner-MULTI CPU & AMD GPU Miner 0.9.0 Win64 & Linux
339915db8f2c15580ac5093ed41bf853 *SRBMiner-Multi-0-9-0-win64.zip
7a6ffbba928423720c0794053ad4f9e2 *SRBMiner-Multi-0-9-0-Linux.tar.xz
- Added algorithm 'blake3_alephium' (ALPH - Alephium coin) for CPU/GPU mining, fee 1.00%
- Added algorithm 'xdag' (XDAG - Dagger coin) for CPU mining, fee 1.00%
- Fixed broken 'yespower' algorithms (broke in previous 0.8.9 version)
- Rewrote miner restarting mechanism on Windows
- Removed algorithm 'rx2'
- Bug fixes
You can use --worker parameter to set worker name when mining 'xdag' algorithm
Because Alephium doesn't have a unified stratum implementation, pools created their own implementations.
I added Herominers and WoolyPooly's implementation, other pools are not guaranteed to work as expected.
Herominers should be preferred because it uses the least network traffic between pool-miner.
You can 'dual' mine 'blake3_alephium' with any other algorithm you like.
Good combinations are : ethash + alph or etchash + alph.
You will need to adapt --gpu-intensity values manually for best results.
SRBMiner-MULTI CPU & AMD GPU Miner 0.8.9 Win64 & Linux
7075e7099a36105ebb5cae7ade894ff0 *SRBMiner-Multi-0-8-9-win64.zip
5fe4a201a38c5e9ffcdb602c1bd08d12 *SRBMiner-Multi-0-8-9-Linux.tar.xz
- Added algorithm 'randoml' (LOZZ - Lozzax coin) for CPU mining, fee 0.85%
- Performance increase on 'ghostrider' algorithm with dynamic thread management
- Performance increase on 'verushash' algorithm for CPU's with AES
- Performance increase on 'scryptn2' algorithm
- MSR tweaks are now allowed for every CPU mineable algorithm
- Dataset for 'autolykos2' algorithm is now created differently, so hopefully video driver won't time-out on slower gpu's
- Added file Help\Tdr_Fix\tdr_fix.reg for Windows, to increase driver time-out value (helps with DAG/dataset creation)
- Removed algorithm 'eaglesong'
- Removed algorithm 'kadena'
- Removed algorithm 'bl2bsha3'
- Removed algorithm 'phi5'
- Removed algorithm 'cryptonight_cache'
- Removed algorithm 'cryptonight_heavyx'
- Minor bug fixes
Dynamic thread management for 'ghostrider' algorithm works only if you leave it on auto setup (don't set --cpu-threads or --cpu-threads-intensity parameter)
'scryptn2' added values 4, 8, 12, 16 for --cpu-threads-intensity. This parameter sets how many hashes to do per worker thread.
--cpu-threads-intensity parameter can accept values per thread if needed, so you can set different number of hashes per thread for every worker thread!
This can further increase the hashrate, if you find the right settings for your system.
Examples:
--cpu-threads 16 --cpu-threads-intensity 16 (uses 16 CPU threads, every thread does 16 hashes)
--cpu-threads 16 --cpu-threads-intensity 8!8!8!8!8!8!8!8!12!12!12!12!12!12!12!12 (uses 16 CPU threads, threads 0..7 do 8 hashes per thread, threads 8..15 do 12 hashes per thread)
SRBMiner-MULTI CPU & AMD GPU Miner 0.8.8 Win64 & Linux
Note:
I re-uploaded this version after fixing a bug in miner API - didn't want to make a new release just because of that
03ef9a371518ae3dde914c1a3aae80d9 *SRBMiner-Multi-0-8-8-win64.zip
d03c6daa7ebce765f7bcbaedc85c88bc *SRBMiner-Multi-0-8-8-Linux.tar.xz
- Added socks5 proxy support (--proxy parameter)
- Added algorithm 'argon2d_16000' (ADOT - Alterdot coin) for CPU/GPU mining, fee 0.85%
- Performance increase on 'heavyhash' algorithm for Vega GPU's (up to ~13%)
- Performance increase on 'argon2id_chukwa2' algorithm for RDNA/RDNA2 GPU's (up to ~12%) & lower power consumption
- Performance increase on 'argon2d_dynamic' algorithm for RDNA/RDNA2 GPU's (up to ~45%) & lower power consumption
- Removed devfee for 'yespoweric' algorithm
- Removed devfee for 'yespowerlitb' algorithm
- Removed devfee for 'yespowerres' algorithm
- If using '--multi-algorithm-job-mode 3' miner now auto sets --max-no-share-sent to 45 minutes if no other value was set
- Rewrote restarting mechanism on Linux so it shouldn't spawn a new process anymore on miner auto restart
- Minor bug fixes
--proxy parameter
Format : ip:port or user:password@ip:port
Supports: tcp, ssl/tls connections
Tested with socks5 proxies from : https://hideip.me/en/proxy/socks5list
Note:
'argon2d_16000' on old drivers (ex. 18.6.1) might produce invalid shares, please use newer drivers
SRBMiner-MULTI CPU & AMD GPU Miner 0.8.7 Win64 & Linux
QUICKFIX
d952e67b87cade96862b1b7e5ded810d *SRBMiner-Multi-0-8-7-win64.zip
e1f842aa1ca09b1c225384a200560b0a *SRBMiner-Multi-0-8-7-Linux.tar.xz
- Fixed hashrate regression on 'heavyhash' algorithm for Navi cards
- Fixed 'progpow' algorithms that got broken in previous release (0.8.6)
SRBMiner-MULTI CPU & AMD GPU Miner 0.8.6 Win64 & Linux
6c0a107ca644d898f12f676f80cfce7d *SRBMiner-Multi-0-8-6-win64.zip
a2c86362a366b36e2dcf3adccc4a6efb *SRBMiner-Multi-0-8-6-Linux.tar.xz
- Performance increase on 'heavyhash' algorithm for Polaris GPU's up to ~10%
- Performance increase on 'heavyhash' algorithm for VEGA GPU's ~5%
- Fixed 'heavyhash' algorithm for RX6500/6600/6700 GPU's
- Fixed detection of GPU's on some linux systems
Coins using 'heavyhash' algorithm :
SRBMiner-MULTI CPU & AMD GPU Miner 0.8.5 Win64 & Linux
036ecaa431eb2f7a747d0fff3c5642f5 *SRBMiner-Multi-0-8-5-win64.zip
f3f7c44d5051cc08559870adce40fdae *SRBMiner-Multi-0-8-5-Linux.tar.xz
- Performance increase on 'heavyhash' algorithm for VEGA/RDNA/RDNA2 GPU's
- Fixed 'verushash' algorithm 'invalid solution version' error
SRBMiner-MULTI CPU & AMD GPU Miner 0.8.4 Win64 & Linux
d6d4dbaa67a3749b7d376dcc819c9518 *SRBMiner-Multi-0-8-4-win64.zip
e42619388a859e398bae2f3641d768a9 *SRBMiner-Multi-0-8-4-Linux.tar.xz
- Performance increase on 'heavyhash' algorithm for GPU's
- Fixed display of hashrate while doing gpu auto tune process
- Changed GPU temperature to show 'edge' value instead of 'hotspot'
SRBMiner-MULTI CPU & AMD GPU Miner 0.8.3 Win64 & Linux
8c26decce57d95b1c58833e25eb994ae *SRBMiner-Multi-0-8-3-win64.zip
dfe78b08fb7fd2ad77cdfa615964b4ec *SRBMiner-Multi-0-8-3-Linux.tar.xz
- Performance increase on 'heavyhash' algorithm for GPU's (up to ~20% on some cards)
- Fixed issue with recognising some GPU's on newer drivers (broken since v0.8.1)
- Reworked hashrate reporting/stats - now reporting average for 1 min / 1 hr / 6 hr / 12 hr (in API too)
SRBMiner-MULTI CPU & AMD GPU Miner 0.8.2 Win64 & Linux
a97dc0b1e7bc8cea495446da3a46bb60 *SRBMiner-Multi-0-8-2-win64.zip
8ca46ad67c585a0a2d95d44d7ce06344 *SRBMiner-Multi-0-8-2-Linux.tar.xz
- Added possibility to preload dataset/s on 'autolykos2' algorithm with '--gpu-autolykos2-preload' parameter (even on 4Gb cards!)
- Little faster dataset creation on some GPU's
- Job notifications are now displayed in console less frequently, unless extended log is enabled
- Reverted 'autolykos2' kernels for hawaii, tonga and fiji to the ones from ver. 0.7.3
- Bug fixes in algorithm switching mode ('--multi-algorithm-job-mode 3')
Dataset preloading can increase pool-side hashrate! (that means more profit)
Works even on GPU's with 4Gb!
For example if you set to preload 1 dataset that means it will create the dataset for the current and next block.
When the next block arrives, dataset will get loaded, which means GPU can continue working immediately (without creating the dataset).
Additional GPU memory needed.
There are 2 types of preloading :
- If you set preload value to 1 it will constantly keep preloading the next block (preferred for 4-8GB gpu's)
- If you set preload value to > 1 it will preload X number of datasets, and when a block arrives that isn't yet loaded, it will be created so miner can start working on it, and then again X number of datasets will get preloaded.