GPU Miner for ETH
, RVN
, BEAM
, CFX
, ZIL
, ERGO
, AE
nbminer.com & NBMiner_github are the only 2 officially maintained site for publishing information and new releases of NBMiner.
Be aware when you download NBMiner binaries from other sources.
version | Windows | Linux |
---|---|---|
42.3 | NBMiner_42.3_Win.zip | NBMiner_42.3_Linux.tgz |
42.2 | NBMiner_42.2_Win.zip | NBMiner_42.2_Linux.tgz |
41.5 | NBMiner_41.5_Win.zip | NBMiner_41.5_Linux.tgz |
41.3 | NBMiner_41.3_Win.zip | NBMiner_41.3_Linux.tgz |
41.0 | NBMiner_41.0_Win.zip | NBMiner_41.0_Linux.tgz |
40.1 | NBMiner_40.1_Win.zip | NBMiner_40.1_Linux.tgz |
Algorithm | Coin | Compute Capability | Memory (Win7 & Linux) | Memory (Win10) |
---|---|---|---|---|
ethash | ETH | 6.0, 6.1, 7.0, 7.5, 8.0,8.6 | 6GB | 6GB |
etchash | ETC | 6.0, 6.1, 7.0, 7.5, 8.0,8.6 | 4GB | 4GB |
cuckoo_ae | AE | 6.0, 6.1, 7.0, 7.5, 8.0,8.6 | 5GB | 6GB |
kawpow | RVN | 6.0, 6.1, 7.0, 7.5, 8.0,8.6 | 4GB | 4GB |
beamv3 | BEAM | 6.0, 6.1, 7.0, 7.5 | 3GB | 3GB |
octopus | CFX | 6.0, 6.1, 7.0, 7.5, 8.0,8.6 | 5GB | 6GB |
ergo | ERGO | 6.0, 6.1, 7.0, 7.5, 8.0,8.6 | 3GB | 3GB |
nbminer -a algo -o protocol+socket_type://pool_host:pool_port -u wallet_address.worker -p passwd
-h, –help Displays this help.
-v, –version Displays version information.
-c, –config <config file path> Use json format config file rather than cmd line options.
–generate-config <filename> Generate a sample config json file.
-a, –algo <algo> Select mining algorithm
–api <host:port> The endpoint for serving REST API.
-o, –url <url> Mining pool url.
-u, –user <user> User used in Mining pool, wallet address or username.
-o1, –url1 <url> url for backup mining pool 1.
-u1, –user1 <user> username for backup mining pool 1.
-o2, –url2 <url> url for backup mining pool 2.
-u2, –user2 <user> username for backup mining pool 2.
-p, –password <password> password for mining pool
-p1, –password1 <password> password for backup mining pool1
-p2, –password2 <password> password for backup mining pool2
-d, –devices <devices> Specify GPU list to use. Format: “-d 0,1,2,3” to use first 4 GPU.
-i, –intensity <intensities> Comma-separated list of intensities (1 -100).
–strict-ssl Check validity of certificate when use SSL connection.
–proxy Socks5 proxy used to eastablish connection with pool, E.g. 127.0.0.1:1080
–cuckoo-intensity <intensity> Set intensity of cuckoo, cuckaroo, cuckatoo, [1, 12]. Smaller value means higher CPU usage to gain more hashrate. Set to 0 means autumatically adapt. Default: 0.
–temperature-limit, –tl <temp-limit> Set temperature limit of GPU, if exceeds, stop GPU.
–temperature-start, –ts <temp-start> Set cool-down temperature target if GPU is stopped by temperature-limit
, default to <temp-limit> - 5.
–log Generate log file named logs/log_<timestamp>.txt
.
–log-file <filename> Generate custom log file. Note: This option will override --log
.
–log-no-job Set this option to disable ‘New job’ info in console and log file.
–log-cycle <cycle> Set to change the cycle of Summary table show in console and log, in seconds, defaults to 30.
–no-health Do not query device health status. (Do not set if use LHR gpu)
–no-color Do not use colorful output log in cmd line.
–long-format Use ‘yyMMdd HH:mm:ss,zzz’ for log time format.
–verbose Print communication data between miner and pool in log file.
–device-info Print device cuda information.
–device-info-json Print out detailed information for every device in json format.
–fee <fee> Change devfee in percentage, [0-5]. Set to ‘0’ to turn off devfee with lower hashrate. Otherwise, devfee = max(set_value, default_value).
–no-watchdog Disable watchdog process.
–platform <platform> Choose platform,0: NVIDIA+AMD (default), 1: NVIDIA only, 2: AMD only
–share-check <value> If <value> minutes without share, reboot miner, set 0 to disable. Default: 30
–no-interrupt set this option will disable miner interrupting current GPU jobs when a new job coming from pool, will cause less power supply issue, but might lead to a bit higher stale ratio and reject shares.
–enable-igpu AMD igpu is disabled by default, set this option to enable.
–mt, –memory-tweak <mode> Memory timings optimize for Nvidia GDDR5 & GDDR5X gpus. range [1-6]. Higher value equals higher hashrate. Individual value can be set via comma seperated list. Power limit may need to be tuned up to get more hashrate. Higher reject share ratio can happen if mining rig hits high temperature, set lower value of -mt
can reduce reject ratio.
–power-limit, –pl <limit> Set power limit of GPU.
Set PL in watts: -pl 200.
Set PL in percetage of default PowerLimit: -pl 75%
–cclock <clocks> Set core clock in MHz. Set clock offsets: -cclock 100, -cclock -500 (Windows only) Set locked clock: -cclock @1500.
–mclock <clocks> Set memory clock offsets in MHz. (Windows only)
–lock-cv <cv> Set locked core voltages in mV. (Windows only)
–fan <speed> Set GPU fan speed in percentage. (Windows only)
–lhr-mode <mode> LHR unlock mode. Mode 1 better compatibility (default mode), Mode 2 better stability.
Open http://api_host:port/ in your browser to use web monitor.
GET http://api_host:port/api/v1/status
{
"miner": {
"devices": [
{
"accepted_shares": 4,
"core_clock": 1530,
"core_utilization": 100,
"fan": 59,
"hashrate": "61.50 M",
"hashrate2": "0.000 ",
"hashrate2_raw": 0,
"hashrate_raw": 61497708.17088159,
"id": 0,
"info": "NVIDIA GeForce RTX 3070",
"invalid_shares": 0,
"lhr": 0,
"memTemperature": -1,
"mem_clock": 8000,
"mem_utilization": 100,
"pci_bus_id": 1,
"power": 131,
"rejected_shares": 0,
"temperature": 57
},
{
"accepted_shares": 3,
"core_clock": 1035,
"core_utilization": 100,
"fan": 80,
"hashrate": "85.07 M",
"hashrate2": "0.000 ",
"hashrate2_raw": 0,
"hashrate_raw": 85069773.34248555,
"id": 1,
"info": "NVIDIA GeForce RTX 3080 Ti LHR",
"invalid_shares": 0,
"lhr": 74,
"memTemperature": 104,
"mem_clock": 10251,
"mem_utilization": 100,
"pci_bus_id": 48,
"power": 258,
"rejected_shares": 0,
"temperature": 64
}
],
"total_hashrate": "146.6 M",
"total_hashrate2": "0.000 ",
"total_hashrate2_raw": 0,
"total_hashrate_raw": 146567481.51336715,
"total_power_consume": 389
},
"reboot_times": 0,
"start_time": 1637549633,
"stratum": {
"accepted_shares": 7,
"algorithm": "ethash",
"difficulty": "4.295 G",
"dual_mine": false,
"invalid_shares": 0,
"latency": 217,
"pool_hashrate_10m": "217.9 M",
"pool_hashrate_24h": "217.9 M",
"pool_hashrate_4h": "217.9 M",
"rejected_shares": 0,
"url": "asia2.ethermine.org:5555",
"use_ssl": true,
"user": "0x4296116d44a4a7259B52B1A756e19083e675062A"
},
"version": "40.0"
}
etchash
: Add support for NiceHash ETC pool.Compared to v41.5
feature
: ethash
Add a new option --lhr-mode
to select LHR unlock mode.
--lhr-mode 1
is the default mode and is the same as which in v41.5
--lhr-mode 2
if stability issue encountered in mode 1
, for json config, use "lhr-mode": "2"
LHR v3
GPUs can only use mode 1
feature
: ethash
Extend LHR unlocker to work with old driver versions. Starting from Windows v512.95
and Linux v515.x
, LHR unlocker no longer works, make sure not to use these new driver versions.optimize
: ethash
small hashrate improvement on Nvidia GPUs.feature
: Add GDDR6X
memory temp in summary table for Linux.feature
: Add GPU RAM type
and GPU RAM vendor
in log.feature
: Add Nvidia driver versoin in summary table.feature
: ethash
Add eth_submitHashrate
for nicehash (EthereumStratum)
protocol.feature
: If Overclock
applies through nbminer
, reset settings to stock when miner exit.feature
: -mt
option on windows does not require custom driver anymore (admin priviledge is required).NOTE
: NBMiner versions older than v39.5 (not included) will run into issue when ETH reaches EPOCH 520
(Approx mid September). Please make sure to upgrade to a newer version before that if ETH is still with POW.feature
: ethash
Add 90% LHR unlocker for 3080 12G
& 3050
.feature
: ergo
Add LHR unlocker support.fix
: ethash
LHR unlocker improve stability.fix
: ethash
Improve stability of LHR unlocker.fix
: ethash
Fix crash on AMD GPUsfix
: ethash
Improve compatibility on rigs with small system memory.note
: Recommend driver versions: 512.15
for Windows, 510.60
for Linux.feature
: ethash
100% LHR unlocker added, for both Windows & Linux.
512.15
for Windows, 510.60
for Linux.fix
: ethash
v40.0 false detected as LHR lock under lastest Nvidia driver versions.feature
: supports future LHR GPU models.feature
: Add overclock options for Nvidia GPUs, require admin priviledge, accept comma seprated list to specify for each GPU.
-power-limit, -pl
: Set power limitation of GPU. Examples:
Set PL in watts: -pl 200
Set PL in percentage of default PowerLimit: -pl 75%
(in Windows bat file, need dual %
, -pl 75%%
)
-cclock
: Set core clock in MHz. Examples:
Set clock offsets: -cclock 100
(Windows only)
Set locked clocks: -cclock @1500
-mclock
: Set memory clock offsets in MHz (Windows only)
-lock-cv
: Set locked core voltage of GPU in mV, support Turing and newer GPUs. (Windows only)
-fan
: Set fan speed in percentage of GPU. (Windows only)
feature
: Display current LHR value in console summary table and web monitor.
feature
: -proxy
options now support username & password for SOCKS5 proxy. format: -proxy user:pass@host:port
feature
: Web monitor changes, delete unused information for dual mining.
feature
: new options for log control.
-log-no-job
: Turn off the New job
line in console.-log-cycle
: Set to change the cycle of Summary table show in console and log, in seconds, defaults to 30.fix
: ethash
Change CPU share validation to independent thread, lower performance degradation when mining at very low difficulty pools.
fix
: LHR lock detected failure on some cases.
feature
: LHR mode support new GA104 version of 3060feature
: ethash
Turn on LHR mode by default for 3060 v1, disabled by default only on windows 470.05fix
: ethash
Fail to detect LHR lock on certain situation.fix
: Nvidia GPU power consumption issue on certain rig config.delete
: support for sero
.optimize
:ethash
LHR mode significant improvement:
-lhr-mode 1
and -lhr-mode 2
-lhr-mode 1
for LHR GPUs-lhr-mode 1
-> 74, -lhr-mode 2
-> 71-lhr-reduce-value
: the amount to reduce -lhr
value on a single -lhr
tuning. defaults to 0.5
.-lhr-reduce-time
: When LHR lock is detected, and the time since the last lock exceeds this value, the -lhr
reduce will not perform. defaults to 15
, which means 15 minutes.-lhr-reduce-limit
: the maximum number of times to reduce -lhr
value, defaults to 6
.feature
: ergo
also adapts the 3 new LHR options.feature
: add -cmd-output
option to specify command line outpu to stdout
or stderr
, 1=stdout
, 2=stderr
, defaults to 2
.feature
: disable SNI extension for ssl connections by default, can be enabled with -enable-sni
optiondelete
: cuckatoo
& cuckatoo32
supportfeature
: ergo
new LHR mode for mining ERGO, enable it by manually adding -lhr
option
ethash
, -lhr
value represents expected hashrate to reach value
percent of non-LHR GPU’s hashrate, supports comma-seperated list to indicate -lhr
value for each GPU, and -lhr -1
to disable.-lhr 85
-lhr 100
-lhr
value by 0.5, and continue mining. max decrease times is 10, which sums to 5.0optimize
: ergo
Lower power consumption on Nvidia GPUs.fix
: octopus
Fix error hash issue.feature
: ethash
new low power LHR mode, add -lhr-mode
option.
-lhr-mode 2
is the default LHR mode, which is the new lower power mode.-lhr-mode 1
changes LHR mode to old version, which is the same as v39.2
-lhr-mode 1
is suitable for only power limit bounded GPU, can achieve higher hashrate than mode 2-lhr-mode 2
is able to achieve lower average power and temperature. espacially suitable for GPUs with gddr6x e.g.3070ti, 3080, 3080ti. Power consumtion is fluctuating in this mode, better be used with locked core clock.feature
: -lhr
support decimal valuefeature
: for LHR GPUs, when mining lock is detected, miner will automatically decrease -lhr
value by 0.1, and continue mining. max decrease times is 10, which sums to 1.0fix
: higher CPU usage when set --share-check 0
feature
: ethash
added LHR lock detection and recovery in LHR mode.feature
: added memory temperature
display (if available) for Nvidia and AMD GPU under windows.fix
: ergo
more robust protocol handle.Note: LHR mode requires NVML library to work, please make sure the driver is installed properly and do not add -no-health
argument.
optimize
: ethash
improve hashrate of LHR
mode 1 - 2%,default value of -lhr
changes from 68 to 69,manually set to 70 is also very promising.fix
: LHR
mode support on windows driver 471.11fix
: kawpow
issue of v39.0fix
: Radeon vii
issue on windows driver 21.6.1+feature
: ethash
New LHR mode for ETH mining on RTX 30 series LHR GPUs, supports Windows & Linux, able to get ~70% of maximum unlocked hashrate.
-lhr
, only works for ethash
right now.-lhr
default to 0, meaning even if -lhr
is not set, LHR mode with -lhr 68
will be applied to LHR GPUs if certain GPUs are detected.-lhr <value>
, a specific value will tell miner try to reach value
percent of maximum unlocker hashrate, e.g. -lhr 68
will expect to get 68% of hashrate for same model non-LHR GPU.-lhr
value will results in higher hashrate, but has higher possibility to run into lock state, which will leads to much less hashrate.-lhr
value can be set for each GPU by using comma separeted list, -lhr 65,68,0,-1
, where -1
means turn off LHR mode.fix
: AMD kernel error on certain GPUs with 21.6.1+ drivers under windows.fix
: high CPU usage in v38.0feature
: ergo
add mining.extranonce.subscribe
support.fix
: eliminate memory leak by NVML library for Nvidia driver 460+ on windows.fix
: --enable-dag-cache
causes crash on certain situation.fix
: ethash
--enable-dag-cache
cause crash on AMD GPUs when switch DAG file.fix
: ergo
support on AMD Vega
GPUs. changes from 37.3
new algo
: ergo
for AMD GPU, can be faster with ETH mining timingsoptimize
: ergo
slightly improce hashrate on Nvidia GPUsfeature
: use --temperature-limit
& --temperature-start
to protect GPU from overheat, detail in readme.mdfeature
: add option --enable-dag-cache
to allow an extra DAG
for different epoch
cached in GPU memory, useful for ETH+ZIL
mining and mining on NiceHash
.feature
: add option -p -p1 -p2
for setting password
of mining pool, old format -u wallet.worker:passwd
is disabled,:
can be added as part of worker or walletoptimize
: ethash
minor hashrate improvement on RDNA GPUsfix
: compatibility issue on lasted AMD 21.4.1
driver.fix
: ergo
high reject ratio on 10 series Nvidia GPUsfix
: ergo
pool compatibilityp106-90
& 1060 3G
to mine ERGO
, hashrate will be increased significantly with -mt
option.new algo
: ergo
for mining ERGO
coin on Nvidia GPUs.delete algo
: bfc
cuckarood
for Nvidia, octopus
for AMDfix
: octopus
support CFX
new address formatfix
: ‘clBuildProgram error’ issue on Vega for versions 35.0 - 36.1feature
: disable AMD iGPU by default, can be enabled back by setting --enable-igpu
other
: minor bug fix, improve overall stabilityoptimize
: octopus
Lower power comsumption for 20、30 series Nvidia GPU, improve hashrate 2% on 16 series Nvidia GPUfix
: ethash
Fix performance degradation under win8 & win8.1 for Nvidia 10 series GPUs.fix
: ethash
Slightly reduce stale ratio.fix
: A random crash bug fix, improve overall stabilityfeature
: Add detail datetime
& cpu usage
in summary logfeature
: ethash
If DAG verification failed, display corresponding GPU name in red in summary.fix
: kawpow
crash on some GPUs in versions 35.x
fix
: ethash
performance degradation under win7 for Nvidia 10 series GPUs.delete algo
: eaglesong
, eaglesong_ethash
, trb
, trb_ethash
, hns
, hns_ethash
, sipc
, cuckaroo
, cuckaroo_swap
feature
: smaller binary sizeCompared to v35.0
optimize
: ethash
More stable under high OC for Nvidia 16、20、30 series GPUs.optimize
: octopus
Improve hashrate 1-3% for 16, 20, 30 Nvidia GPUsfeature
: ethash
DAG verification after creation, if miner showed log in red font: Verification failed, invalid 2.0%
, please consider lower GPU overclockoptimize
: ethash
More stable under high OC for Nvidia GPUsoptimize
: octopus
Improve hashrate 1-3% for 16, 20, 30 Nvidia GPUsfeature
: ethash
Add statistics for invalid shares
, in cmd log, api and web monitor.feature
: Turn off limitation for not allowing to run under Virtual Machine.feature
: Add statistics for Health information of AMD GPUfix
: More detail error information of OpenCL apifix
: Reduce CPU usage.optimize
: ethash
Improve hashrate 1% on certain Nvidia GPUsoptimize
: octopus
Minor improvement on certain 20 & 30 series Nvidia GPUsfeature
: -mt
More effective and compatibility on Geforce Pascal GPUsfix
: kawpow
progpow_sero
Fix crash on certain AMD & Nvidia rigsoptimize
: octopus
Improve hashrate 1-5% on Nvidia 16, 20, 30 series GPUs, 29.2M on 1660sfix
: etchash
error on swiching epoch, ETC
miners should upgrade to this version.fix
: support for AMD 20.11.x driver versionfix
: ethash
more stable hashrate under windowsfix
: ethash
Reduce stale share ratio.fix
: ethash
Fix display hashrate only half of normal hashrate on certain windows rigs.optimize
: ethash
Improve hashrate on Nvidia 10 series GPUs,3% higher hashrate under same PowerLimit, or same hashrate with 5%-10% lower PowerLimit.feature
: octopus
Add support for mining on NiceHash
feature
: Add an option -no-interrupt
, set this option will disable miner interrupting current GPU jobs when a new job coming from pool, will cause less power supply issue, but might lead to a bit higher stale ratio and reject shares.feature
: Add effiecieny
display in console, showing hashrate per watt
for each GPUfeature
: Add 10min 4h 24h pool hashrate display in web monitor.optimize
: octopus
improve hashrate: +10% on 16 20 30 series Nvidia GPUs, 27.5M on 1660s.note
: octopus
From this version, GPUs that has higher core performance than memory performance, need to overclock memory to get higher hashrate, e.g. 2080 3070optimize
: octopus
improve hashrate: +90% on 16 20 30 series Nvidia GPUs, at least +100% on all other GPUsoptimize
: octopus
improve hashrate: +35% on 16 20 30 series Nvidia GPUs, +20% on all other GPUsnew algo
: etchash
for upcoming ETC upgradefeature
: add effective pool hashrate on console & api, 10min 4h 24h.optimize
: octopus
improve hashrate: Nvidia, +3% on 10 series, +20% on 16, 20, 30 seriesnew algo
: octopus
for AMD support.fix
: crash upon start on certain Nvidia rigs.optimize
: octopus
improve hashrate: +150% on 10 series, +80% on 16, 20, 30 seriesfix
: --share-check 0
cause high CPU usagefix
: Added back Uptime
in console summaryfix
: some format error on console printoptimize
: octopus
lower CPU usagefeature
: add new option --share-check
, if no share found in a set period of time, miner will reboot. default to 30 minutes. SOLO miners should set this option to 0
to turn off check.new algo
: octopus
for miningconflux
,support both solo mining and pool mining, need Nvidia GPU above 6Goptimize
: ethash
improve performance on Vega & Navi GPUsoptimize
: beamv3
improve performance on high end 10xx Nvidia GPUsfeature
: modify summary output on console, add share statistics for each GPU.fix
: ethash
fix zero hashrate on certain cases for AMD GPUsbug fix
: Fix AMD device initialization failure on some rigs.new algo
: beamv3
for mining BEAM
with Nvidia 3GB+ GPUsnew algo
: cuckatoo32
add support for Nvidia 6GB GPUoptimize
: ethash
for AMD RX 4xx, 5xx, Vega series 8GB+ GPUsfeature
: ethash
support mining up to epoch 800new algo
: cuckatoo32
for Grin32
on Nvidia 8G above GPUskawpow
: support for NiceHash
’s extranonce.subscribe
protocolwindows
: Auto install driver if --memory-tweak
is set and driver is not installed.windows
: Fix driver install failed
issue on some windows rigs.kawpow
: Fix duplicate share
issue on some pools.ethash
kawpow
startup time.invalid kernel image
.-mt
is set to 0 on corresponding gpu.--memory-tweak
, optimize memory timings of Nvidia GD5 & GD5X GPUs. Detail describe can be found in readme.md--verbose
, print pool communucation log.--proxy
, user can using socks5 proxy to set up connection with pool.kawpow
on AMD RDNA GPU.kawpow
mining on Nvidia & AMD gpus.miningrigrentals
.-di
calculation is changed in this version for HNS+ETH.
-di
= work_size_of_hns
/ work_size_of_eth
-di 5
on a stock freq 1070ti will get 26M for eth & 130M for hns.--platform
to allow users to choose GPU platform.CKB
mining on AMD cards under linux.CKB
mining on AMD cards.CKB
mining & CKB
+ETH
dual mining.-di
parameter to support comma separated list to specify -di
value for each card.-di
for CKB
+ETH
mining ranges from 4 ~ 8 depending on GPU model, valid value range in [1, 10], higher value means higher intensity for ETH
.-di
differs by GPU model, overclock and power limitation.OhGodAnETHlargementPill
to boost SIPC performance.--generate-config
to generate a sample config file.start_sero.sh
cuckaroo
, v24.0 will do an automatically switch to cuckarood
on height 262080.-grin29-fork-height
, user can test the auto-switch by setting this option to lower height value.SERO
, algo progpow_sero
.-intensity
to set the intensity level for each GPU.-fidelity-timeframe
to customize the timeframe for fidelity calculation.-log-file
to set a specified log filename.-no-nvml
to close the periodical query for GPU status.nicehash+tcp://
--cuckoo-intensity
to 4, lower CPU usage on default settings when mining Grin & AE."GPU hung detected"
happens under some normal conditions.-di
for dual mining BTM+ETH.ethnh
)BTM+ETH
dual mining mode.