============================ NiceHash QuickMiner v0.5.3.3 ============================ Instructions: 1. Start NiceHashQuickMiner.exe. 2. You will be asked for your NiceHash mining address. Provide it. 3. All further interaction with NiceHash QuickMiner is through notification icon in taskbar. Right click it and then perform any action you want. 4. It is recommended to set autostart with Windows if this is your classis rig machine. 5. Use OCTune to adjust overclocks and fan speeds (do not use Afterburner, it is not needed). 6. Enjoy carefree mining! serviceLocation -1 = auto serviceLocation 0 = eu west serviceLocation 1 = usa west serviceLocation 2 = eu north serviceLocation 3 = usa east Change serviceLocation, BTC, workerName, launchCommandLine. consoleLogLevel, fileLogLevel (for NiceHashQuickMiner); use following numbers: 0-6 which means: 0: log everything (TRACE) 1: log debug (DEBUG) 2: standard (INFO) 3: warnings (WARNING) 4: errors (ERROR) 5: fatal (FATAL) 6: no logging (disabled) Please, enable full file logging ("fileLogLevel":0) and submit log files when error is found. Updating from previous version? Make sure to full uninstall it first - remove Windows autostart in old version before you enable autostart in new one. From version 0.2.0.0, CPU Mining with integrated XMRig is possible on modern CPUs with AVX2. CPU Mining can be activated from the tray menu (or via Rig Manager). You can provide extra launch parameters for XMRig (config nhqm.conf), for example: "CPUMinerELP":"--cpu-max-threads-hint 50 --print-time 15" First parameter above tells CPU miner to use only 50% of resources, second one prints speed to console every 15 seconds. Parameters for lowest CPU priority, donation level 0% and other important parameters are already provided by the Excavator. There are 3 options what NiceHashQuickMiner should do when device fails or when device has too high speed. "whenDeviceSpeedTooHigh":2, "whenDeviceSpeedZero":1, 1 = no action 2 = restart Excavator 3 = restart rig Usually, when device has too high speed, issue can be resolved by restarting Excavator. If device has speed 0 (errored out), it means that there was some instability (either too high OC or issue with risers). In that case, it is most likely better to restart the rig (set to value 3). From version 0.3.0.0, NiceHashQuickMiner has auto updater. You will be notified when new version is available for download. You can participate in testing RC versions (release candidate) which have new features faster; enable this by setting: "bUpdateRCVersion":true Latest Excavator has two more events: on_quickminer.start and on_quickminer.stop. You can put setting OC profile for mining there so when you stop the miner, mining OC profile is reset back to your regular OC settings and you can play a game. After you return back and click start again, mining OC profile will be applied. At the moment, this has to be done manually. First find your optimal OC with OCTune and save it. Then open commands.json file and find your GPU overclock command which should look like this: { "id": 1, "method": "device.set.oc_profile", "params": ["0", "-400", "900", "122"] } Move (cut it out, including {} brackets) and put it into event "on_quickminer.start". If this event does not exist in your commands.json file, you can add it or rename existing "on_quit" event. You also need to add OC reset command when mining stops. Add following command into "on_quickminer.stop": { "id": 1, "method": "device.set.oc_reset", "params": ["0"] } After modification, your commands.json file content should be something like that (assuming your gaming device ID is 0): ... }, { "event": "on_quit", "commands": [] }, { "event": "on_quickminer.start", "commands": [{ "id": 1, "method": "device.set.oc_profile", "params": ["0", "-400", "900", "122"] }] }, { "event": "on_quickminer.stop", "commands": [{ "id": 1, "method": "device.set.oc_reset", "params": ["0"] }] } ] There are two methods to apply OC profile. Alternative one takes only max core clock and absolute memory clock. This means you can reach your ideal OC configuration faster, because first, you need to determine your max. memory stability clock and after that just decrease max core clock so you get wanted efficiency - leave it a bit higher for higher speed and drop it down if you want less power consumption. RTX 3060 Ti wants around 1350 MHz, RTX 3070 around 1200 MHz and for 3090 120 MH/s is possible with 290 W (clock limit 1095 MHz) which means 412 kH/J efficiency. With 0.3.0.5 experimental autotuner has been added. You can now determine almost perfect OC settings for your card for max efficiency with just one mouse click in 5 minute time! Feature is available under OCTune. Version 0.3.0.7 brings GDDR6X temperature measurements (only RTX 3080 and RTX 3090). Version 0.3.1.0 brings optimize profiles (two profiles which can be applied without struggling for doing any manual OC). These profiles are applied remotely over Rig Manager. If you have your own OC, then you have to have profile "Manual" selected or your OC settings may get overwritten! Version 0.3.1.1 uses new Excavator which has a smartfan functionality - can set fan speed according to temperature of your GPU or VRAM and GPU. Old fan management is deprecated and replaced with smartfan. There are following smartfan modes: 1. No FAN management by Excavator (mode = 0) 2. Fixed speed (mode = 1) 3. Target GPU temperature; fan speed is set according to wanted GPU temp and current GPU temp (mode = 2) 4. Target GPU&VRAM temperature; as above, but also VRAM temp is considered (mode = 3) OCTune has new UI which support these changes so fan management is easier now. Also fan settings can be saved just like OC settings and are applied with each Excavator (re)start. Version 0.3.1.2 contains bug fixes. Service autostart method removed. Task scheduler method is much better and gives less problems. Excavator's speed for Pascal cards is restored back to what was in version 0.3.0.2. Version 0.3.2.0 has improved speed measuring. If you are upgrading, make sure to delete periodic method "workers.reset.all" from your commands.json file. If you are not sure how to do it properly, just start from zero. There are some other less important bug fixes and improvements. 0.3.2.0 comes with installer & uninstaller (Windows Add/Remove Programs). Version 0.3.2.1 contains mostly just bug fixes and some small improvements when launching for the first time. Version 0.3.2.2 greatly improves SmartFan. Obtaining mining domains for correct function of NiceHash QuickMiner is now done using DNS over HTTPS which should reduce number of issues users having with certain Firewalls and other security solutions which are unjustifiably blocking traffic to NiceHash. Handling of erroneous state has been improved (rig is not restarted if there is no internet and first it tries to restart Excavator). Also because we consider this as a major stability configuration, this setting has been moved into firstrun optional settings dialog. 0.3.2.3 adds "Game Mode". This mode disables mining on primary display GPU, applies gaming overclock and allows you to switch into gaming mode with just a click of a button. When you are done, it is possible to return into mining mode just as easily. Note that this feature is experimental. Some other bugs were also fixed. 0.3.2.5 and 0.3.2.6 mostly just bug fixes and small improvements. 0.3.2.7 brings back power readings for Mobile versions of GPUs. 0.3.3.0 has automatic location select option which is recommended to be set by everyone. Automatic location works by first testing latency of accepted shares for all possible mining locations (currently there are 4). It remembers this data and keep it for 7 days - after 7 days, this test is performed again because network conditions may change over time. Additionally, when you have automatic location selected, if NiceHash informs QuickMiner that certain location is not available, QuickMiner will select next best location by latency automatically thus keep your Rig mining without any outtage. So it is strongly recommended to have automatic location selected. If you already have certain location selected, you can modify nhqm.conf and set -1 for "serviceLocation". 0.3.4.0 also measures additional temperatures (HotSpot - VRAM) for other GPUs besides RTX 3080 and 3090 and gives you ability to adjust fans accordingly. 0.4.0.0 with GPU core undervoltage support. Works on all supported cards. Please, be careful when using this feature. We do not allow setting manual voltage of more than 1000 mV, but we cannnot guarantee that there is no bug in the code that would somehow allow to bypass this limit and consequently cause damage to your GPU. Only use if you know what you are doing! 4th error recovery mode was added - restart of NVIDIA driver. 0.4.1.1, 0.4.1.2 and 0.4.1.3 are mostly small improvements and bug fixes. 0.4.1.4 addresses issue with altered names of devices when using NVIDIA developer drivers 470.05 and caused OPTIMIZE not working. From this version on, SmartFan gets reset only if being set to anything but auto. This happens when Excavator exits automatically so devices.smartfan.reset is not needed for event "on_quit" in commands.json anymore. This would make less problems when using 3rd party tools to manage your fans. 0.4.2.0 brings ability to set memory timings. Currently only two timings which are needed for GDDR5X to achieve good performance. Set FAW to 16 and RRD to 4. If that does not work, you can try FAW to 20 and RRD to 5. Example: {"id":1,"method":"device.set.memory.timings","params":["0","FAW=16","RRD=4"]} If you wish to reset memory timings, set memory delta clock to 0. When DAG is being generated, memory timings are reset to default and then reapplied (smart DAG gen just like mem OC has). It is recommended to start this command with some delay to avoid instant crash. 0.4.3.0 brings OPTIMIZE for certain Pascal cards (1080 Ti and 1080 initially). From version 0.4.4.0 more various OPTIMIZE profiles are available and new ones added modified daily. Be sure to periodically check it out! These get better and better over time. In 0.4.5.0 access to HTTP API was restructured to not allow CORS anymore. It was pointed out by one security researcher that a remote change of mining address could be possible if user visits a malicious website. To remedy this security bug, OCTune is now loaded by Excavator's HTTP server and CORS is not turned on anymore. Also an 'Authorization' token is set and all API calls need it to be able to make any changes to the mining process. There are also some further improvements regarding CUDA detection and ID assignments which may further reduce issues with Invalid DAG file generated problems. 0.4.5.3 is another security focused upgrade which brings N/M signature support for update and install packages. Using several signatures to execute upgrade or install greatly increases resiliency towards malicious attempts to distribute malware through update packages in case there was a security breach at NiceHash. v0.4.5.4 and v0.4.5.5 just bug fixes and small improvements. v0.5.0.0; Game Mode: if you would like to override auto detected display device, then fill "overrideDisplayDeviceUUID" in your nhqm.conf with the UUID value of your GPU. You can get UUID value of your GPU if you stop/disable device in Rig Manager and then check your config to see which UUID was disabled. v0.5.0.0 also adds MSI-enable tweak: all your supported GPUs are set to use MSI. This can significantly improve stability of your system. GPU crashes are way easier to handle then and in most cases can be handled with simple driver restart (or restart of Excavator). You can read more about MSI here: https://en.wikipedia.org/wiki/Message_Signaled_Interrupts By default, NVIDIA drivers does not use MSI for GTX 1000 and RTX 2000 series cards, which is a big mistery because Pascal and Turing architecture does support MSI. Every time drivers are updated, this value is changed back to not using MSI, so you need to apply it after each driver update. Target plus decrease MCC fan mode (version 0.5.1.0+) is an inteligent Smart Fan mode which works similar as Target Core&VRAM (mode 3) with upgrade - when max set fan speed is reached and temperature is still over target, MCC (max core clock) is being reduced thus reducing power output of the GPU which most likely reduces temperature. The newest Smart Fan mode takes in one more parameter - max fan speed which is identical to Smart Fan's advanced parameters = override_level_max (default = -1 = use GPU's default max level = in most cases this is 100%). This fan mode enables you to really take control of the temperatures and make sure to keep them in check with drastic measures (reducing clocks, load and power) if needed. It is NOT available for Pascal architecture, but fully supported for Volta, Turing and Ampere, because prior enabling it, alternative OC needs to be set with defined MCC (core clock limit). With version 0.5.2.0 new nhqm.conf variables were added: * environmentVariables: string array of EnvironmentVariables for Excavator.exe (eg: "CUDA_VISIBLE_DEVICES=0,1", "CUDA_LAUNCH_BLOCKING=1", ...) * bDisableConfigFileChecks: disable reading of nhqm.conf file during working (may improve issues with corrupted config file)