Antminer Vnish Firmware Guide, Screen by Screen

Real screenshots from a Vnish 1.3.5 miner on our bench, not mock-ups. The menu on the left and the tabs along the top work exactly as they do on the machine: click them in the screenshot and this guide follows. Blue boxes move you to another screen, orange boxes explain a control — what it does, what a healthy value looks like, and what actually goes wrong when it is set badly. All nineteen screens are here: the dashboard and every log tab, the mining settings, and the system side including flashing. Pool addresses, worker names and serial numbers are blurred.

The same explanations as the interactive guide above, written out so they can be read, searched and quoted without clicking anything.

Dashboard and the panels around it

Every tile on these four screens, in text

Dashboard

Pool addresses and worker names are blurred. Everything else is exactly as the miner shows it.

Vnish 1.3.5 dashboard on an Antminer S21 Pro: hashrate, temperature, PSU and error tiles above the hashrate and temperature charts, the pool table and the three hashboard cards
DashboardMonitoring › Dashboard The monitoring screen. Nothing here changes anything - it is everything the miner is doing right now, on one page.Where you should be looking most of the time.It is read-only, which also means a problem visible here has to be fixed on one of the other screens. Do not go hunting for a fix on the dashboard.
MiningSettings › Mining Four tabs: Pools, Cooling, Performance, Advanced. Everything that decides how the machine runs lives behind this one menu item.Visited deliberately, not browsed.Three of these four tabs can cost you hardware. Advanced says so itself at the top of the page.
SystemSettings › System Four tabs: General, Security, Network, Maintenance. Housekeeping, plus the two buttons that undo everything.Password changed from the default, timezone correct, a backup taken.On 1.3.5 backup, restore, reset and rollback sit on Maintenance. The vendor's own System documentation still describes three tabs and puts them under General, so it will send you to the wrong place.
Tech supportTech support Opens a panel with four tiles: a diagnostic report, the API documentation the miner serves itself, a Telegram channel that is not live yet, and the vendor's documentation site.The diagnostic report is the one that earns its place.For a hardware fault this is the wrong door. A board that will not detect chips is not a firmware ticket - generate a diagnostic report - from here or from the Logs panel toolbar - and send it to us instead.
Firmware updateFirmware update Opens the flashing dialog over whatever page you were on.Opened when you actually intend to flash.Nothing happens until a file is dropped in and Flash miner is pressed - but that is the only safety net between here and a rewritten control board.
Quick actionsthe toggle button in the header Opens the panel with interface language, light or dark theme, optimised rendering and the Locate miner button.Whatever suits you - none of it changes how the machine mines.Locate miner is the exception. It makes the miner blink so you can pick it out of a row of identical boxes, which is a physical action on a running machine.
Info panelthe panel button in the header Slides out a column with events from the last 24 hours, the firmware and hardware identity, network status and memory use.Where you read off the model, the build, the control board platform and the address without leaving the page.This is the fastest place to confirm which platform you actually have before you download a firmware file for it.
Power menuthe orange power button Opens a short menu: throttle mode, pause, restart mining, reboot the device.Closed. You open it, pick one thing, it closes.Restart mining and Reboot device are two different sizes of hammer. The first re-runs the mining process; the second reboots the control board and costs you the cool-down and start-up again.
HashrateHashrate tile Three numbers, not one. The big figure is what the miner is doing now. Curr is the short average, Theor is what the current preset and frequency say it should be doing.Curr within about 3% of Theor. Here that is 234.3 against 236.8.A gap wider than roughly 5% that does not close means something is wrong further down: errors climbing, a chain dropping, or a preset the cooling cannot hold. Look at the board cards before you touch a setting.
TemperatureTemperature tile The big range is chip temperature. The small line underneath is the circuit board. Chip temperature is what the protections act on.Chips comfortably below the critical value set in Advanced, which defaults to 90 °C on this build.One hot chip drags the top of the range up long before the average looks bad. The firmware watches the hottest chip, not the mean - a 20° spread inside one board is worth investigating.
PSUPSU tile Draw in watts, plus the internal temperatures of the two hot stages inside the power supply: the PFC stage and the LLC converter.Watts tracking the preset. A 3320 W preset pulling about 3570 W at the wall is normal - the preset names the chip budget, not the whole machine.PSU internal temperatures climbing while the room stays the same is a supply that is aging or starved of air. That failure takes hashboards with it when it goes.
ErrorsErrors tile The share of computed work the pool will not accept, as a percentage and as an absolute hashrate.At or near zero.Anything sustained above roughly half a percent means unstable frequency, heat, or a board on its way out. Errors move earlier than hashrate, which averages the problem away.
Performance and efficiencyPerformance tile Active preset, whether it has finished tuning, autoswitch state, efficiency in joules per terahash, and the PSU model the firmware detected.A tuned badge. Until it says tuned, the numbers beside it are provisional.Vnish prints efficiency in J/TH here even though the vendor documentation says J/GH. Trust the screen.
Hashrate chartHashrate graph Hashrate over time with power on the second axis. Range switches between 1 h, 12 h and 24 h. The menu exports SVG, PNG or CSV.A flat line with narrow noise.The vertical drops to zero on this chart are restarts. Regular ones you did not cause are the single most useful thing to notice on this screen.
Temperature chartTemperature graph Chip temperature over time with fan duty on the second axis.Temperature flat, fan duty with headroom left.Temperature steady only because fan duty keeps climbing is a machine running out of cooling. The chart shows that; the tile does not.
Pool tablePools One row per pool. Diff is share difficulty, Accepted and Rejected are counts, Stale is work that arrived too late, Last share is time since the pool took anything.One pool marked Active with a latency figure, the rest Alive and idle. Last share resetting every few seconds.Last share creeping into minutes while hashrate looks fine means you are computing into a hole. Rejected climbing steadily is usually latency or an overclock the pool keeps refusing.
DevFee indicatorDevFee Shows the developer fee connection is up. The fee is taken as redirected hashrate while the miner runs - 3% on this build - not as a bill.Online.It is disclosed rather than hidden on purpose. Firmware advertised as fee-free either hides the fee somewhere you cannot see it, or is not what it claims to be.
Cooling summaryCooling Live RPM for every fan, plus the mode, the chip temperature being held, and current duty.All fans reporting, RPM within a few hundred of each other, duty below 100% if the room allows.Duty pinned at 100% while temperature still sits at target means no headroom left - the next warm day costs hashrate. A fan showing no RPM is a fan the miner will eventually stop for.
Hashboard cardBoard 1 / 2 / 3 Per board: its own hashrate and how close that is to expected (HR %), errors, temperature range, frequency, voltage, consumption, and a status line.65 - All chips are good and HR above about 95%, with the three boards within a few percent of each other.Any status line other than all chips good is a board to look at, not a setting to change. Failed to detect ASIC chips at 0.00 H/s is a dead chain - the miner keeps running on the others, but that board is a repair.
Logs panelLogs Opens along the bottom without leaving the page. Five tabs: Status, Miner, HWscan, Autotune, System.Quiet.HWscan is not documented by the vendor anywhere. Before you send a miner in, generate a diagnostic report from here - it saves a day of back and forth.
Status barbottom right Model, control board platform and firmware version, always visible.A platform badge matching the file you flashed - aml here for Amlogic.This is the fastest way to confirm which board family you have before downloading anything. If it says xil and you are about to flash an aml build, stop.

Power menu

The yellow button in the top right. Everything that starts, stops or restarts the miner lives behind it.

Vnish 1.3.5 power menu on an Antminer S21 Pro with throttle mode, pause, restart mining and reboot device
DashboardMonitoring › Dashboard The monitoring screen. Nothing here changes anything - it is everything the miner is doing right now, on one page.Where you should be looking most of the time.It is read-only, which also means a problem visible here has to be fixed on one of the other screens. Do not go hunting for a fix on the dashboard.
MiningSettings › Mining Four tabs: Pools, Cooling, Performance, Advanced. Everything that decides how the machine runs lives behind this one menu item.Visited deliberately, not browsed.Three of these four tabs can cost you hardware. Advanced says so itself at the top of the page.
SystemSettings › System Four tabs: General, Security, Network, Maintenance. Housekeeping, plus the two buttons that undo everything.Password changed from the default, timezone correct, a backup taken.On 1.3.5 backup, restore, reset and rollback sit on Maintenance. The vendor's own System documentation still describes three tabs and puts them under General, so it will send you to the wrong place.
Tech supportTech support Opens a panel with four tiles: a diagnostic report, the API documentation the miner serves itself, a Telegram channel that is not live yet, and the vendor's documentation site.The diagnostic report is the one that earns its place.For a hardware fault this is the wrong door. A board that will not detect chips is not a firmware ticket - generate a diagnostic report - from here or from the Logs panel toolbar - and send it to us instead.
Firmware updateFirmware update Opens the flashing dialog over whatever page you were on.Opened when you actually intend to flash.Nothing happens until a file is dropped in and Flash miner is pressed - but that is the only safety net between here and a rewritten control board.
The yellow buttontop right of the header Opens this menu. It is the only place in the interface that can stop a running miner.Closed, most of the time.Two of the four items here take the machine off the pool immediately, with no confirmation step worth the name. Read before you click.
Throttle modeThrottle mode - Adjustable power operation Scales the miner's output by a percentage without changing the preset or stopping anything. Set it to 50% and a 230 TH machine does roughly 115 TH.100%, which is what the status bar shows during normal running.It is the polite way to shed load on a hot afternoon or a constrained circuit. Left at 50% and forgotten, it looks exactly like a miner that has quietly lost half its hashrate.
PausedPaused - Stop mining Stops mining but leaves the firmware and the interface running, so the machine stays reachable.Used when you are about to work on the machine.A paused miner earns nothing and looks healthy from the outside - fans turning, interface responding, dashboard up. Check the status bar at the bottom right before you conclude a miner is fine.
Restart miningRestart mining Stops and restarts the mining process only. The control board does not reboot.The first thing to try when hashrate has drifted and nothing obvious explains it.It clears the symptom without telling you the cause. If you find yourself doing it weekly, the answer is in the Logs panel, not in this button.
Reboot deviceReboot device Full restart of the control board, as if you had cut the power and put it back.Rarely needed. Restart mining first.Takes minutes rather than seconds, and on a machine with a marginal power supply a reboot is exactly the moment things fail to come back. Do not reboot remotely if nobody can get to the machine.

Quick actions

The small button left of the power menu. Interface preferences plus one genuinely useful field tool.

Vnish 1.3.5 quick actions menu with interface language, UI theme, optimized rendering and the locate miner function
The buttontop right, left of the power menu Opens the Quick actions panel. Everything in it is either cosmetic or physical - nothing here changes how the miner mines.Safe to poke at.The only item with a real-world effect is Locate miner, and even that only flashes a light.
Interface languageInterface language Same list as System › General: English, Russian, Ukrainian, Portuguese, Persian and Georgian.Whatever you read fastest.If you are sending us screenshots, English saves us translating your interface before we can read your problem.
UI themeUI theme Light, dark, or OS Sync which follows your computer.Whichever you can read on a phone in a shed.Purely cosmetic.
Optimized renderingOptimized rendering Disables animation and interface effects.Enabled if the interface feels sluggish, especially on a phone.Changes nothing about mining. It only makes the page lighter to draw.
Locate minerLocate miner Flashes the LEDs on the physical machine so you can find it in a rack.The fastest way to match an IP address to a box in a room full of identical boxes.Genuinely useful and completely harmless. Remember to look at the machine while it is flashing - it does not run forever.

Info panel

Slides out from the header button next to the power menu. PSU serial and MAC address are blurred here.

Vnish 1.3.5 info panel on an Antminer S21 Pro showing system information, network status, memory use and events in the last 24 hours
PinPin Keeps the panel open instead of letting it slide away.Pinned while you are working on the machine.Nothing breaks. It just closes on you mid-read if you leave it unpinned.
Events in 24hEvents in 24h Counter and list of notable events over the last day, with the miner's own clock and timezone shown underneath.Zero.The timezone link here goes straight to System › General. If the clock is wrong, every timestamp you ever send us is wrong with it.
System InformationSystem Information The identity of the machine: model, stock hashrate, firmware version, build date, build name and UUID, control board platform, install type, PSU model and serial.Platform and Installation Type matching the firmware file you flashed - aml and NAND here.This block is the answer to almost every question we ask before a repair. Screenshot it and send it with the request. Build name also tells you whose build you are running: ours reads VnishnetMRepair.
Network statusNetwork status Hostname, DHCP or static, IP, netmask, gateway, both DNS entries and the MAC address, read-only. The arrow jumps to System › Network to change any of it.An address you can reach and a hostname that identifies this machine.Read-only here on purpose. The place things go wrong is the Network tab, where a wrong gateway makes the miner unreachable.
MemoryMemory RAM in use on the control board.Around a third. Ours sits at 32% of 244 Mb.Memory climbing steadily over days on a machine you have not touched is a control board worth watching. It is not a setting - it is a symptom.

The five log tabs

What each log tab prints, and what it is telling you

Logs: Status

Opens along the bottom from the Logs bar, over whatever page you were on.

Vnish 1.3.5 Status log opened along the bottom of the miner interface over the page you were on
DashboardMonitoring › Dashboard The monitoring screen. Nothing here changes anything - it is everything the miner is doing right now, on one page.Where you should be looking most of the time.It is read-only, which also means a problem visible here has to be fixed on one of the other screens. Do not go hunting for a fix on the dashboard.
MiningSettings › Mining Four tabs: Pools, Cooling, Performance, Advanced. Everything that decides how the machine runs lives behind this one menu item.Visited deliberately, not browsed.Three of these four tabs can cost you hardware. Advanced says so itself at the top of the page.
SystemSettings › System Four tabs: General, Security, Network, Maintenance. Housekeeping, plus the two buttons that undo everything.Password changed from the default, timezone correct, a backup taken.On 1.3.5 backup, restore, reset and rollback sit on Maintenance. The vendor's own System documentation still describes three tabs and puts them under General, so it will send you to the wrong place.
Tech supportTech support Opens a panel with four tiles: a diagnostic report, the API documentation the miner serves itself, a Telegram channel that is not live yet, and the vendor's documentation site.The diagnostic report is the one that earns its place.For a hardware fault this is the wrong door. A board that will not detect chips is not a firmware ticket - generate a diagnostic report - from here or from the Logs panel toolbar - and send it to us instead.
Firmware updateFirmware update Opens the flashing dialog over whatever page you were on.Opened when you actually intend to flash.Nothing happens until a file is dropped in and Flash miner is pressed - but that is the only safety net between here and a rewritten control board.
Status tabLogs › Status The plain narrative: cooling down, auto-tuning, start mining, initialising.Cycles at times you can explain.Restarts you did not schedule are the whole point of this tab.
Miner tabLogs › Miner The mining process in detail: chip detection, voltages, preheating, frequency ramp, pool activity.Every chain reporting its full chip count at start-up.First tab to read when hashrate is wrong and the dashboard looks normal.
HWscan tabLogs › HWscan The hardware scan: what the firmware found on each chain and on the PSU at start-up.All chains connected, all EEPROMs read.Undocumented by the vendor anywhere, and the tab that tells you a board came up short before the hashrate does.
Autotune tabLogs › Autotune Progress and results of tuning runs, per chain, as percentages of expected.A run that ends with Auto-tune finished successfully.Tuning that never finishes means the machine cannot hold the preset. That is cooling or power, not firmware.
System tabLogs › System Kernel-level output: filesystem, network interface, serial ports.Boring, and ending with the link up.Unreadable unless you already know what you are looking for - but it is where a control board with failing storage shows itself first.
What Status showslog output Timestamped lines, newest at the bottom. Ours shows the daily cycle: Initializing, Cooling down the miner, Cooling down completed, Start mining.A rhythm you recognise.Ours restarts once a day at 01:00. If you see a pattern like that and cannot say what causes it, find out before it turns into a pattern you cannot ignore.
Panel toolbartop right of the Logs panel Copy, generate a diagnostic report, clear the log, collapse and close.The report button is the one that matters.Clearing the log throws away the history of the fault you are trying to diagnose. Generate the report first - it packages all five tabs plus board data, temperatures and configuration into one file.

Logs: Miner

The most useful tab on a machine that is misbehaving.

Vnish 1.3.5 Miner log showing chip detection, voltage and preheating and the frequency ramp on an Antminer S21 Pro
DashboardMonitoring › Dashboard The monitoring screen. Nothing here changes anything - it is everything the miner is doing right now, on one page.Where you should be looking most of the time.It is read-only, which also means a problem visible here has to be fixed on one of the other screens. Do not go hunting for a fix on the dashboard.
MiningSettings › Mining Four tabs: Pools, Cooling, Performance, Advanced. Everything that decides how the machine runs lives behind this one menu item.Visited deliberately, not browsed.Three of these four tabs can cost you hardware. Advanced says so itself at the top of the page.
SystemSettings › System Four tabs: General, Security, Network, Maintenance. Housekeeping, plus the two buttons that undo everything.Password changed from the default, timezone correct, a backup taken.On 1.3.5 backup, restore, reset and rollback sit on Maintenance. The vendor's own System documentation still describes three tabs and puts them under General, so it will send you to the wrong place.
Tech supportTech support Opens a panel with four tiles: a diagnostic report, the API documentation the miner serves itself, a Telegram channel that is not live yet, and the vendor's documentation site.The diagnostic report is the one that earns its place.For a hardware fault this is the wrong door. A board that will not detect chips is not a firmware ticket - generate a diagnostic report - from here or from the Logs panel toolbar - and send it to us instead.
Firmware updateFirmware update Opens the flashing dialog over whatever page you were on.Opened when you actually intend to flash.Nothing happens until a file is dropped in and Flash miner is pressed - but that is the only safety net between here and a rewritten control board.
Status tabLogs › Status The plain narrative: cooling down, auto-tuning, start mining, initialising.Cycles at times you can explain.Restarts you did not schedule are the whole point of this tab.
Miner tabLogs › Miner The mining process in detail: chip detection, voltages, preheating, frequency ramp, pool activity.Every chain reporting its full chip count at start-up.First tab to read when hashrate is wrong and the dashboard looks normal.
HWscan tabLogs › HWscan The hardware scan: what the firmware found on each chain and on the PSU at start-up.All chains connected, all EEPROMs read.Undocumented by the vendor anywhere, and the tab that tells you a board came up short before the hashrate does.
Autotune tabLogs › Autotune Progress and results of tuning runs, per chain, as percentages of expected.A run that ends with Auto-tune finished successfully.Tuning that never finishes means the machine cannot hold the preset. That is cooling or power, not firmware.
System tabLogs › System Kernel-level output: filesystem, network interface, serial ports.Boring, and ending with the link up.Unreadable unless you already know what you are looking for - but it is where a control board with failing storage shows itself first.
Chip detectionchain#N - NN chips detected At every start the firmware counts the chips on each chain and prints the result.The same number on every chain - 65 on all three here.A chain that reports fewer chips than its neighbours is a hashboard that has already lost silicon. This line finds it hours before the hashrate average does.
Voltage and preheatingSetting voltage / Preheating chips Applied voltage in millivolts, then the preheat, then the switch from manual to automatic fan control at the target temperature.Preheating completed, then Waiting for first job.A machine that never reaches Preheating completed is failing before it ever mines - look at the chain lines above, not at the pool.
Frequency rampRaising freq ... gradually / chip by chip The firmware walks the frequency up rather than jumping to target, first per chain and then chip by chip.A ramp that reaches the preset frequency and stays.A ramp that stalls or restarts is the machine telling you the preset is beyond what this hardware and this cooling can hold.
Panel toolbartop right of the Logs panel Copy, generate a diagnostic report, clear the log, collapse and close.The report button is the one that matters.Clearing the log throws away the history of the fault you are trying to diagnose. Generate the report first - it packages all five tabs plus board data, temperatures and configuration into one file.

Logs: HWscan

Board and PSU identification at start-up. Serial numbers are blurred here.

Vnish 1.3.5 HWscan log with PSU identification and chain connection and EEPROM reads for each hashboard
DashboardMonitoring › Dashboard The monitoring screen. Nothing here changes anything - it is everything the miner is doing right now, on one page.Where you should be looking most of the time.It is read-only, which also means a problem visible here has to be fixed on one of the other screens. Do not go hunting for a fix on the dashboard.
MiningSettings › Mining Four tabs: Pools, Cooling, Performance, Advanced. Everything that decides how the machine runs lives behind this one menu item.Visited deliberately, not browsed.Three of these four tabs can cost you hardware. Advanced says so itself at the top of the page.
SystemSettings › System Four tabs: General, Security, Network, Maintenance. Housekeeping, plus the two buttons that undo everything.Password changed from the default, timezone correct, a backup taken.On 1.3.5 backup, restore, reset and rollback sit on Maintenance. The vendor's own System documentation still describes three tabs and puts them under General, so it will send you to the wrong place.
Tech supportTech support Opens a panel with four tiles: a diagnostic report, the API documentation the miner serves itself, a Telegram channel that is not live yet, and the vendor's documentation site.The diagnostic report is the one that earns its place.For a hardware fault this is the wrong door. A board that will not detect chips is not a firmware ticket - generate a diagnostic report - from here or from the Logs panel toolbar - and send it to us instead.
Firmware updateFirmware update Opens the flashing dialog over whatever page you were on.Opened when you actually intend to flash.Nothing happens until a file is dropped in and Flash miner is pressed - but that is the only safety net between here and a rewritten control board.
Status tabLogs › Status The plain narrative: cooling down, auto-tuning, start mining, initialising.Cycles at times you can explain.Restarts you did not schedule are the whole point of this tab.
Miner tabLogs › Miner The mining process in detail: chip detection, voltages, preheating, frequency ramp, pool activity.Every chain reporting its full chip count at start-up.First tab to read when hashrate is wrong and the dashboard looks normal.
HWscan tabLogs › HWscan The hardware scan: what the firmware found on each chain and on the PSU at start-up.All chains connected, all EEPROMs read.Undocumented by the vendor anywhere, and the tab that tells you a board came up short before the hashrate does.
Autotune tabLogs › Autotune Progress and results of tuning runs, per chain, as percentages of expected.A run that ends with Auto-tune finished successfully.Tuning that never finishes means the machine cannot hold the preset. That is cooling or power, not firmware.
System tabLogs › System Kernel-level output: filesystem, network interface, serial ports.Boring, and ending with the link up.Unreadable unless you already know what you are looking for - but it is where a control board with failing storage shows itself first.
PSU identificationDetecting board / PSU model / PSU fw version Control board platform, then the power supply model, firmware version, serial and history status.A PSU model the firmware recognises and a clean history status.PSU history status other than zeroes means the supply has recorded faults. That is a supply on its way out, and it will take hashboards with it.
Chain connection and EEPROMchain#N - connected / Reading EEPROM's Each chain answers, then the firmware reads the EEPROM that holds each board's identity and configuration.All chains connected, EEPROMs read without complaint.A chain that does not connect, or an EEPROM that will not read, is a hardware fault. No firmware setting fixes either - that is an EEPROM repair or a board repair.
Per-chain identitychain#N - machine, bin, freq, volt, hashrate, sn For each board: the machine code, its bin, the factory frequency and voltage, expected hashrate and the board serial.Three boards with the same machine code and similar bins.bin is the factory silicon grade. A board with a very different bin from its neighbours will always tune lower, and no amount of frequency will change that.
The 2015 timestamps[2015/01/01 00:00:18] The scan runs before the clock is set, so these lines carry the control board's default date rather than the real one.Normal. Nothing is wrong.Do not use HWscan timestamps to correlate with anything. Use the Status or Miner tabs, which run after the clock syncs.
Panel toolbartop right of the Logs panel Copy, generate a diagnostic report, clear the log, collapse and close.The report button is the one that matters.Clearing the log throws away the history of the fault you are trying to diagnose. Generate the report first - it packages all five tabs plus board data, temperatures and configuration into one file.

Logs: Autotune

What tuning actually did, chain by chain, as a percentage of expected.

Vnish 1.3.5 Autotune log showing a tuning run finishing chain by chain as a percentage of expected hashrate
DashboardMonitoring › Dashboard The monitoring screen. Nothing here changes anything - it is everything the miner is doing right now, on one page.Where you should be looking most of the time.It is read-only, which also means a problem visible here has to be fixed on one of the other screens. Do not go hunting for a fix on the dashboard.
MiningSettings › Mining Four tabs: Pools, Cooling, Performance, Advanced. Everything that decides how the machine runs lives behind this one menu item.Visited deliberately, not browsed.Three of these four tabs can cost you hardware. Advanced says so itself at the top of the page.
SystemSettings › System Four tabs: General, Security, Network, Maintenance. Housekeeping, plus the two buttons that undo everything.Password changed from the default, timezone correct, a backup taken.On 1.3.5 backup, restore, reset and rollback sit on Maintenance. The vendor's own System documentation still describes three tabs and puts them under General, so it will send you to the wrong place.
Tech supportTech support Opens a panel with four tiles: a diagnostic report, the API documentation the miner serves itself, a Telegram channel that is not live yet, and the vendor's documentation site.The diagnostic report is the one that earns its place.For a hardware fault this is the wrong door. A board that will not detect chips is not a firmware ticket - generate a diagnostic report - from here or from the Logs panel toolbar - and send it to us instead.
Firmware updateFirmware update Opens the flashing dialog over whatever page you were on.Opened when you actually intend to flash.Nothing happens until a file is dropped in and Flash miner is pressed - but that is the only safety net between here and a rewritten control board.
Status tabLogs › Status The plain narrative: cooling down, auto-tuning, start mining, initialising.Cycles at times you can explain.Restarts you did not schedule are the whole point of this tab.
Miner tabLogs › Miner The mining process in detail: chip detection, voltages, preheating, frequency ramp, pool activity.Every chain reporting its full chip count at start-up.First tab to read when hashrate is wrong and the dashboard looks normal.
HWscan tabLogs › HWscan The hardware scan: what the firmware found on each chain and on the PSU at start-up.All chains connected, all EEPROMs read.Undocumented by the vendor anywhere, and the tab that tells you a board came up short before the hashrate does.
Autotune tabLogs › Autotune Progress and results of tuning runs, per chain, as percentages of expected.A run that ends with Auto-tune finished successfully.Tuning that never finishes means the machine cannot hold the preset. That is cooling or power, not firmware.
System tabLogs › System Kernel-level output: filesystem, network interface, serial ports.Boring, and ending with the link up.Unreadable unless you already know what you are looking for - but it is where a control board with failing storage shows itself first.
The tuning runchain#N - NNNNN GH/s (NN.NN%) Each pass records the voltage and frequency it tried and what every chain returned, in GH/s and as a percentage of what that chain should manage.Percentages in the high nineties and close to each other. Ours sit around 96 to 98%.One chain persistently ten points below the others is a weak board, not a tuning problem. That is the board to pull when you next have the machine open.
Finish lineAuto-tune finished successfully The line that says the run completed and the result was stored.Present.If you never see this line, the tune did not finish. Anything you concluded from the hashrate in the meantime was measured mid-tune and means nothing.
Panel toolbartop right of the Logs panel Copy, generate a diagnostic report, clear the log, collapse and close.The report button is the one that matters.Clearing the log throws away the history of the fault you are trying to diagnose. Generate the report first - it packages all five tabs plus board data, temperatures and configuration into one file.

Logs: System

Kernel output from the control board. Mostly for us, not for you.

Vnish 1.3.5 System log with filesystem, ethernet and serial port messages from the Antminer control board
DashboardMonitoring › Dashboard The monitoring screen. Nothing here changes anything - it is everything the miner is doing right now, on one page.Where you should be looking most of the time.It is read-only, which also means a problem visible here has to be fixed on one of the other screens. Do not go hunting for a fix on the dashboard.
MiningSettings › Mining Four tabs: Pools, Cooling, Performance, Advanced. Everything that decides how the machine runs lives behind this one menu item.Visited deliberately, not browsed.Three of these four tabs can cost you hardware. Advanced says so itself at the top of the page.
SystemSettings › System Four tabs: General, Security, Network, Maintenance. Housekeeping, plus the two buttons that undo everything.Password changed from the default, timezone correct, a backup taken.On 1.3.5 backup, restore, reset and rollback sit on Maintenance. The vendor's own System documentation still describes three tabs and puts them under General, so it will send you to the wrong place.
Tech supportTech support Opens a panel with four tiles: a diagnostic report, the API documentation the miner serves itself, a Telegram channel that is not live yet, and the vendor's documentation site.The diagnostic report is the one that earns its place.For a hardware fault this is the wrong door. A board that will not detect chips is not a firmware ticket - generate a diagnostic report - from here or from the Logs panel toolbar - and send it to us instead.
Firmware updateFirmware update Opens the flashing dialog over whatever page you were on.Opened when you actually intend to flash.Nothing happens until a file is dropped in and Flash miner is pressed - but that is the only safety net between here and a rewritten control board.
Status tabLogs › Status The plain narrative: cooling down, auto-tuning, start mining, initialising.Cycles at times you can explain.Restarts you did not schedule are the whole point of this tab.
Miner tabLogs › Miner The mining process in detail: chip detection, voltages, preheating, frequency ramp, pool activity.Every chain reporting its full chip count at start-up.First tab to read when hashrate is wrong and the dashboard looks normal.
HWscan tabLogs › HWscan The hardware scan: what the firmware found on each chain and on the PSU at start-up.All chains connected, all EEPROMs read.Undocumented by the vendor anywhere, and the tab that tells you a board came up short before the hashrate does.
Autotune tabLogs › Autotune Progress and results of tuning runs, per chain, as percentages of expected.A run that ends with Auto-tune finished successfully.Tuning that never finishes means the machine cannot hold the preset. That is cooling or power, not firmware.
System tabLogs › System Kernel-level output: filesystem, network interface, serial ports.Boring, and ending with the link up.Unreadable unless you already know what you are looking for - but it is where a control board with failing storage shows itself first.
FilesystemUBIFS lines The control board mounting its own storage at boot.Mounts cleanly and moves on.Repeated errors here are failing flash on the control board. That is not a firmware problem and reflashing will not fix it.
Ethernetmeson6-dwmac ... eth0 The network interface coming up, with the negotiated link speed.Link is Up - 100Mbps/Full.A link that flaps, or negotiates at 10 Mbps, is a cable or a port - and it looks exactly like a pool problem from the dashboard.
Serial portsmeson_uart ... ttyS The serial channels the control board uses to talk to the hashboards, changing baud rate as the chains come up.Steady progression to the working rate.This is diagnostic depth you will never need unless we ask for it. Send the report rather than reading it yourself.
Panel toolbartop right of the Logs panel Copy, generate a diagnostic report, clear the log, collapse and close.The report button is the one that matters.Clearing the log throws away the history of the fault you are trying to diagnose. Generate the report first - it packages all five tabs plus board data, temperatures and configuration into one file.

Mining settings

Every control on the mining tabs, in text

Pools

URL and worker fields are blurred - on your own miner they identify you to the pool.

Vnish 1.3.5 Pools tab on an Antminer S21 Pro with the pool slots in priority order and the URL, worker and password fields
DashboardMonitoring › Dashboard The monitoring screen. Nothing here changes anything - it is everything the miner is doing right now, on one page.Where you should be looking most of the time.It is read-only, which also means a problem visible here has to be fixed on one of the other screens. Do not go hunting for a fix on the dashboard.
MiningSettings › Mining Four tabs: Pools, Cooling, Performance, Advanced. Everything that decides how the machine runs lives behind this one menu item.Visited deliberately, not browsed.Three of these four tabs can cost you hardware. Advanced says so itself at the top of the page.
SystemSettings › System Four tabs: General, Security, Network, Maintenance. Housekeeping, plus the two buttons that undo everything.Password changed from the default, timezone correct, a backup taken.On 1.3.5 backup, restore, reset and rollback sit on Maintenance. The vendor's own System documentation still describes three tabs and puts them under General, so it will send you to the wrong place.
Tech supportTech support Opens a panel with four tiles: a diagnostic report, the API documentation the miner serves itself, a Telegram channel that is not live yet, and the vendor's documentation site.The diagnostic report is the one that earns its place.For a hardware fault this is the wrong door. A board that will not detect chips is not a firmware ticket - generate a diagnostic report - from here or from the Logs panel toolbar - and send it to us instead.
Firmware updateFirmware update Opens the flashing dialog over whatever page you were on.Opened when you actually intend to flash.Nothing happens until a file is dropped in and Flash miner is pressed - but that is the only safety net between here and a rewritten control board.
Pools tabMining › Pools Where the computed work is sent. Three slots, priority by order.A main pool and a backup on a different provider.The only tab in Mining that cannot damage hardware. It can still cost you a day of mining into nothing.
Cooling tabMining › Cooling Fan mode, how many fans are required, the temperature to hold, and the duty range.Auto mode, full duty range available.Every setting here trades heat for noise or hashrate. Heat is the one that ends in a repair.
Performance tabMining › Performance Presets, autotuning, global voltage and frequency, per-board offsets and the per-chip map.A preset showing tuned and offsets the tuner chose itself.This is where hand-editing undoes a three-hour tune in one click. Change the preset, not the individual numbers.
Advanced tabMining › Advanced Protections, restart behaviour, power limits, and a section the firmware itself labels Danger zone.Left at defaults unless you know precisely why not.The firmware prints Any changes made in this section are at your own risk at the top. It means it.
Priority is the orderinfo banner The miner's own wording: priority is controlled by the order in the list, and you change it by dragging.Your main pool first, a backup on a different provider second.Three slots on one provider is not redundancy. One outage and you have three dead pools.
Drag handlePool: 1 Grab here to move a pool up or down the list.Deliberate order you can explain.Reordering is instant and silent. It is easy to demote your paying pool by accident and not notice for a day.
Pool URLURL Full stratum address including the port.Exactly as your pool documents it.A typo here looks identical to a network fault from the dashboard: the pool sits at Alive, never goes Active, and Last share never moves.
WorkerWorker Your pool account plus a name for this specific machine, usually joined with a dot.A name you will recognise in the pool dashboard at 3 a.m.Naming every machine the same thing means the pool shows you one blob of hashrate and you lose the ability to spot which unit dropped.
PasswordPassword Most pools ignore this or expect a placeholder such as x. Some use it to carry options.Whatever your pool's documentation says.Not a security field. Do not put anything here you would mind reading in a support screenshot.

Cooling

Four settings. Three are safe to touch. One is not.

Vnish 1.3.5 Cooling tab with cooling mode, minimum number of fans, target chip temperature and the fan speed limit
DashboardMonitoring › Dashboard The monitoring screen. Nothing here changes anything - it is everything the miner is doing right now, on one page.Where you should be looking most of the time.It is read-only, which also means a problem visible here has to be fixed on one of the other screens. Do not go hunting for a fix on the dashboard.
MiningSettings › Mining Four tabs: Pools, Cooling, Performance, Advanced. Everything that decides how the machine runs lives behind this one menu item.Visited deliberately, not browsed.Three of these four tabs can cost you hardware. Advanced says so itself at the top of the page.
SystemSettings › System Four tabs: General, Security, Network, Maintenance. Housekeeping, plus the two buttons that undo everything.Password changed from the default, timezone correct, a backup taken.On 1.3.5 backup, restore, reset and rollback sit on Maintenance. The vendor's own System documentation still describes three tabs and puts them under General, so it will send you to the wrong place.
Tech supportTech support Opens a panel with four tiles: a diagnostic report, the API documentation the miner serves itself, a Telegram channel that is not live yet, and the vendor's documentation site.The diagnostic report is the one that earns its place.For a hardware fault this is the wrong door. A board that will not detect chips is not a firmware ticket - generate a diagnostic report - from here or from the Logs panel toolbar - and send it to us instead.
Firmware updateFirmware update Opens the flashing dialog over whatever page you were on.Opened when you actually intend to flash.Nothing happens until a file is dropped in and Flash miner is pressed - but that is the only safety net between here and a rewritten control board.
Pools tabMining › Pools Where the computed work is sent. Three slots, priority by order.A main pool and a backup on a different provider.The only tab in Mining that cannot damage hardware. It can still cost you a day of mining into nothing.
Cooling tabMining › Cooling Fan mode, how many fans are required, the temperature to hold, and the duty range.Auto mode, full duty range available.Every setting here trades heat for noise or hashrate. Heat is the one that ends in a repair.
Performance tabMining › Performance Presets, autotuning, global voltage and frequency, per-board offsets and the per-chip map.A preset showing tuned and offsets the tuner chose itself.This is where hand-editing undoes a three-hour tune in one click. Change the preset, not the individual numbers.
Advanced tabMining › Advanced Protections, restart behaviour, power limits, and a section the firmware itself labels Danger zone.Left at defaults unless you know precisely why not.The firmware prints Any changes made in this section are at your own risk at the top. It means it.
Cooling modeCooling mode Auto holds a target temperature. Manual pins the fans at a fixed percentage. Immersion turns fan handling off for liquid setups.Auto, for almost everyone.Manual on an air-cooled miner means the fans stop responding to a hot day. Immersion mode on an air-cooled miner means no fan supervision at all - that is how machines cook.
Minimum number of fansMinimum number of fans How many working fans the miner insists on before it will run.The number actually fitted. Four on a standard air-cooled unit.Lowering this to keep a machine running with a dead fan is a decision to run hot deliberately. It works right up until it does not, and the damage lands on a hashboard.
Target temperatureTarget temperature The chip temperature auto mode aims to hold by moving the fans.65 °C, which is the shipped default here.Raising the target to quieten the fans raises every chip temperature by the same amount. Chip life falls off sharply with sustained heat and you will not see the cost for months.
Limit fan speedLimit fan speed A floor and a ceiling on fan duty. Auto mode works inside this range.A wide range. 10 to 100% here.Capping the ceiling below 100% for noise removes the headroom the miner needs on the hottest day of the year - which is exactly the day it will need it.

Performance

The grid under each board is the per-chip hashrate map. Pale numbers are chips the tuner marked weak.

Vnish 1.3.5 Performance tab with the autotune preset, global voltage and frequency, per-board offsets and the per-chip hashrate map
DashboardMonitoring › Dashboard The monitoring screen. Nothing here changes anything - it is everything the miner is doing right now, on one page.Where you should be looking most of the time.It is read-only, which also means a problem visible here has to be fixed on one of the other screens. Do not go hunting for a fix on the dashboard.
MiningSettings › Mining Four tabs: Pools, Cooling, Performance, Advanced. Everything that decides how the machine runs lives behind this one menu item.Visited deliberately, not browsed.Three of these four tabs can cost you hardware. Advanced says so itself at the top of the page.
SystemSettings › System Four tabs: General, Security, Network, Maintenance. Housekeeping, plus the two buttons that undo everything.Password changed from the default, timezone correct, a backup taken.On 1.3.5 backup, restore, reset and rollback sit on Maintenance. The vendor's own System documentation still describes three tabs and puts them under General, so it will send you to the wrong place.
Tech supportTech support Opens a panel with four tiles: a diagnostic report, the API documentation the miner serves itself, a Telegram channel that is not live yet, and the vendor's documentation site.The diagnostic report is the one that earns its place.For a hardware fault this is the wrong door. A board that will not detect chips is not a firmware ticket - generate a diagnostic report - from here or from the Logs panel toolbar - and send it to us instead.
Firmware updateFirmware update Opens the flashing dialog over whatever page you were on.Opened when you actually intend to flash.Nothing happens until a file is dropped in and Flash miner is pressed - but that is the only safety net between here and a rewritten control board.
Pools tabMining › Pools Where the computed work is sent. Three slots, priority by order.A main pool and a backup on a different provider.The only tab in Mining that cannot damage hardware. It can still cost you a day of mining into nothing.
Cooling tabMining › Cooling Fan mode, how many fans are required, the temperature to hold, and the duty range.Auto mode, full duty range available.Every setting here trades heat for noise or hashrate. Heat is the one that ends in a repair.
Performance tabMining › Performance Presets, autotuning, global voltage and frequency, per-board offsets and the per-chip map.A preset showing tuned and offsets the tuner chose itself.This is where hand-editing undoes a three-hour tune in one click. Change the preset, not the individual numbers.
Advanced tabMining › Advanced Protections, restart behaviour, power limits, and a section the firmware itself labels Danger zone.Left at defaults unless you know precisely why not.The firmware prints Any changes made in this section are at your own risk at the top. It means it.
PresetPreset The operating profile. The name is a power budget and the hashrate it is expected to produce. Applying one restarts mining and starts autotuning.A preset the cooling can actually hold, showing tuned.Presets were reworked in the 1.3 branch for the S21 family, S19 XP+ and L9. If you updated from 1.2.x on those, retune rather than assuming the old profile carried across.
Autoswitch presetsAutoswitch presets Lets the firmware move between presets by itself, based on chip temperature, power draw and stability.Off is a fine default. On, with bounds set, suits a site whose temperature swings across the day.Switched on without bounds it chases conditions and you lose the ability to reason about what the machine is doing. Set the ceiling and floor first.
Global voltageGLOBAL VOLTAGE Chip voltage for the whole miner inside the active preset. The line underneath shows the nominal value and how far you are from it.At or slightly below nominal, set by the tuner rather than by hand. Here 13.5 V against a nominal 13.8 V.Voltage is the fastest way to kill chips. Up buys stability now and costs lifespan later. RESET TO NOMINAL is right there for when you have lost track.
Global frequencyGLOBAL FREQUENCY Base clock for the whole miner. Individual boards and chips sit above or below it.Whatever the autotuner settled on. Here 580 MHz against a nominal 595.Raising this by hand after a successful tune undoes the tune. If you want more hashrate, pick a bigger preset and let it retune.
Per-board offsetboard badge Each hashboard carries its own offset from the global frequency. The tuner sets these to even out real silicon differences.Offsets in both directions. Here +52, -34 and +53 MHz across three boards, which is a normal spread.A board sitting far below the others is not one to push up. It is the weakest silicon in the machine and the tuner already knows.
Per-chip hashrate mapHashrate - GH/s Every ASIC on the board with its own measured hashrate. This is the view stock firmware simply does not have.A tight spread with a handful of pale entries.Pale values are chips the tuner de-emphasised. A cluster of them in one physical area of a board is a heat or power-delivery problem there, and it is the single most useful screenshot you can send us with a repair.
Parameter analysis timePerformance › Autoswitch · not on the capture above How long autoswitch watches temperature, power draw and stability before it decides to move a preset. Vendor docs recommend five minutes and allow one to ten.Five minutes.Too short and it reacts to a passing spike, shuffling presets for no reason. Too long and it is slow to back off a machine that is genuinely overheating.
Don’t raise preset above / Don’t lower preset belowPerformance › Autoswitch · not on the capture above The two limits autoswitch may not cross. The switch temperatures themselves are yours to set in the firmware, not fixed by it — the ceiling keeps it out of a profile your cooling cannot hold, the floor keeps it off a hashrate not worth the power.Set both before switching autoswitch on.Without a floor a machine with a real fault winds itself down to almost nothing and keeps drawing power while it does. Without a ceiling it climbs until the cooling gives out.
Exceeding nominal consumptionPerformance · not on the capture above Watches actual draw against what the profile expects, set as a percentage or in watts. Past the limit the firmware lowers the profile by itself.On, with a margin matched to your circuit rather than to the miner’s nameplate.This one protects the breaker, not the miner. Without it a profile drawing more than planned keeps drawing until something upstream trips.

Advanced

Control at the top, Others and the Danger zone below - the whole tab in one screen. The warning across the top is the firmware's own wording, not ours.

Vnish 1.3.5 Advanced tab on an Antminer S21 Pro: restart on low hashrate, critical chip temperature, power limit, the Others toggles and the danger zone
DashboardMonitoring › Dashboard The monitoring screen. Nothing here changes anything - it is everything the miner is doing right now, on one page.Where you should be looking most of the time.It is read-only, which also means a problem visible here has to be fixed on one of the other screens. Do not go hunting for a fix on the dashboard.
MiningSettings › Mining Four tabs: Pools, Cooling, Performance, Advanced. Everything that decides how the machine runs lives behind this one menu item.Visited deliberately, not browsed.Three of these four tabs can cost you hardware. Advanced says so itself at the top of the page.
SystemSettings › System Four tabs: General, Security, Network, Maintenance. Housekeeping, plus the two buttons that undo everything.Password changed from the default, timezone correct, a backup taken.On 1.3.5 backup, restore, reset and rollback sit on Maintenance. The vendor's own System documentation still describes three tabs and puts them under General, so it will send you to the wrong place.
Tech supportTech support Opens a panel with four tiles: a diagnostic report, the API documentation the miner serves itself, a Telegram channel that is not live yet, and the vendor's documentation site.The diagnostic report is the one that earns its place.For a hardware fault this is the wrong door. A board that will not detect chips is not a firmware ticket - generate a diagnostic report - from here or from the Logs panel toolbar - and send it to us instead.
Firmware updateFirmware update Opens the flashing dialog over whatever page you were on.Opened when you actually intend to flash.Nothing happens until a file is dropped in and Flash miner is pressed - but that is the only safety net between here and a rewritten control board.
Pools tabMining › Pools Where the computed work is sent. Three slots, priority by order.A main pool and a backup on a different provider.The only tab in Mining that cannot damage hardware. It can still cost you a day of mining into nothing.
Cooling tabMining › Cooling Fan mode, how many fans are required, the temperature to hold, and the duty range.Auto mode, full duty range available.Every setting here trades heat for noise or hashrate. Heat is the one that ends in a repair.
Performance tabMining › Performance Presets, autotuning, global voltage and frequency, per-board offsets and the per-chip map.A preset showing tuned and offsets the tuner chose itself.This is where hand-editing undoes a three-hour tune in one click. Change the preset, not the individual numbers.
Advanced tabMining › Advanced Protections, restart behaviour, power limits, and a section the firmware itself labels Danger zone.Left at defaults unless you know precisely why not.The firmware prints Any changes made in this section are at your own risk at the top. It means it.
The warningAny changes made in this section are at your own risk The firmware puts this banner here itself.Read as written.Everything below can be set to a value that runs the machine into damage. None of it is validated against your cooling.
Restart on low hashrateRestart on low hashrate Restarts the miner automatically if hashrate drops below a percentage of what it should be doing.Off, or a genuinely low threshold. Ships disabled.Set too close to normal and the miner restarts every time the pool hiccups, losing more work than the dip did. Restart loops also hide the real fault.
Chips critical temperatureChips critical temperature Mining stops when any chip passes this. The firmware watches the hottest chip, not the average.90 °C, the shipped default on this build.This is the last line of defence between a hot spot and a dead board. Raising it to stop a machine cutting out does not fix the heat, it removes what was protecting you from it. This build ships it at 90.
Max restart attemptsMax restart attempts How many consecutive automatic restarts before the miner gives up and stays down.3. Enough to ride out a blip, few enough to stop a loop.0 means infinite. A miner with a real hardware fault will restart forever, heating and cooling the boards every cycle - worse for them than simply stopping.
Minimum number of operational boardsMinimum number of operational boards The miner stops if fewer active chains than this are working. This is the setting that lets a three-board machine keep earning on two.1 to keep running through a board failure. 3 if you would rather it stop and tell you.At 1 a machine can quietly run for weeks on a third of its hashrate and you only notice on the invoice.
Power LimitPower Limit A hard ceiling in watts. Past it the firmware pulls frequency and voltage down by itself. The range for this model is printed in the field.Off unless your circuit, PDU or contract genuinely constrains you.Set below what your preset needs and you have quietly capped your own hashrate - and it looks like the miner underperforming rather than doing what you told it.
Mining startup delay rangeMining startup delay range in seconds A random wait before mining starts, picked from this range.0 for one machine. A spread of a minute or two for a room full of them.Everything on a site restarting at once after a power cut is an inrush spike that trips breakers. This exists to stagger that.
Quick startQuick start Skips some start-up checks to begin mining sooner.Off. The checks are short and they exist.On some models it changes nothing because the cooling countdown is fixed. On others you are trading verification for a minute.
Check balance of voltage domainsCheck balance of voltage domains Watches for uneven voltage distribution across a board and tries to restore stable operation. Vendor docs call this Voltage balancing control.On.Off removes an early warning of a power-delivery fault, which is one of the failures that takes chips with it.
Check integrity of chainsCheck integrity of chains Polls every chip on every board. On miners with a PIC controller it shuts down only the faulty board. On the rest it stops the whole miner — unless Allow disabling chains without PIC in the Danger zone below is switched on, which is what lets a 1.3.5 machine without a PIC keep mining on its healthy chains. Most current hardware falls in that second group: the PIC belongs to the S19 generation and the 21 series ships without one. Vendor docs call this Circuit integrity check.On. Leave it on.Off means a board that has stopped responding keeps drawing power and heating up with nothing watching it.
Adapt voltage to temperatureAdapt voltage to temperature Nudges chip voltage as temperature moves - cold chips often need a little more to stay stable, hot chips need less.On.Off is only sensible if you are hand-tuning in a temperature-controlled room and know exactly why.
Restore miner state on bootRestore miner state on boot After a reboot, put the miner back into whatever state it was in rather than starting fresh.Depends whether you want a paused machine to stay paused through a power cut.With this off, a miner you deliberately paused starts mining again by itself after an outage.
Launch the miner in pause modeLaunch the miner in pause mode Boots into pause instead of mining.Off for production. On for a bench machine you want to inspect before it loads up.Left on by accident, you have a rack that came back from an outage and did nothing.
Quiet modeQuiet mode Holds fans at 50% while the miner is starting or stopping, to cut noise.Off unless the miner lives somewhere people sleep.Slower cooling during those windows. Harmless on a cool site, less so in a hot room.
Downscale preset on failureDownscale preset on failure Drops the preset a level on overheating or chain errors instead of stopping. Vendor docs call this Automatic profile reduction.On is the kinder setting for a machine you cannot reach quickly.It masks the underlying problem while keeping you earning. Fine as a stopgap, bad as a permanent state - find out why it downscaled.
Auto retune on chain breakAuto retune on chain break On a chain-break error the firmware does not just restart - it resets the profile and retunes from scratch.On.Expect a three-to-four hour tune and unstable hashrate after it fires. That is the feature working, not a fault.
Offline miningOffline mining Keeps the miner running with no internet. Nothing is submitted, so nothing is earned - it only holds temperature and load steady.Off for normal operation.Easy to forget it is on and wonder why a machine that looks busy pays nothing.
Ignore non-functional temperature sensorsDanger zone Lets the miner run while ignoring sensors it considers faulty.Off, unless we told you to turn it on for a specific diagnosis.Vendor docs are explicit that only non-critical sensors can be ignored - the critical ones stay under control whatever this is set to. That narrows the risk without removing it: a front sensor reading wrong and a board actually cooking still look the same from here.
Allow disabling chains without PICDanger zone Permits shutting down a board on hardware that has no PIC controller to do it cleanly. Worth knowing which machines that means: the PIC belongs to the S19 generation, and the whole 21 series ships without one — so on most current hardware this is not an exotic switch, it is the only way to keep mining on the boards that still work.Off by default, for a reason.This is how you keep mining on the good boards of a machine whose hardware was never designed to isolate one. It works, and it carries real risk of damage. Worth it during a repair window, not as a permanent configuration.
Tuner bad chip hashrate thresholdTuner bad chip hashrate threshold How far below par a chip has to fall before the tuner treats it as bad and works around it. The default value is printed underneath the slider.The shipped default is 0%.Set it high and the tuner writes off chips that were merely warm, and you lose hashrate you had. At zero it persists with chips that are dragging the chain.
Check voltage settingsAdvanced › Others · not on the capture above, PIC miners only Checks that the voltage actually reaching each chain matches what the profile asked for. Vendor docs call it Voltage control and say it exists only on miners with a PIC controller. In practice that means the S19 generation: the 21 series ships without a PIC, so on most current hardware this control is simply absent — including the S21 Pro these screenshots were taken on.On, wherever the miner offers it.With it off a chain can sit at the wrong voltage while every reading upstream looks normal. Undervolted chips throw errors that read like a tuning fault; overvolted ones age quietly.
Automatic chip regulationAdvanced · not on the capture above Redistributes frequency between chips on a chain so their temperatures even out, instead of letting the hottest chip set the ceiling for the whole board.On, unless you tune per chip by hand and want your own numbers left alone.Off, one warm corner of a board decides when the whole chain throttles, and you pay for it in hashrate everywhere else.
Voltage boost on startupAdvanced › Danger zone · not on the capture above Lifts voltage briefly while the chips initialise, to bring up a chain that otherwise refuses to start.Off. This is a diagnostic aid, not an operating mode.Vendor docs warn against leaving it on: the same extra voltage that starts a stubborn chain will overheat it if it runs that way. A board that only starts with this enabled needs looking at, not a permanent boost.

System, support and flashing

Every control in System, plus support and the firmware update dialog

General

Four tabs live under System: General, Security, Network and Maintenance.

Vnish 1.3.5 System General tab on an Antminer S21 Pro with timezone, interface language, theme and rendering options
DashboardMonitoring › Dashboard The monitoring screen. Nothing here changes anything - it is everything the miner is doing right now, on one page.Where you should be looking most of the time.It is read-only, which also means a problem visible here has to be fixed on one of the other screens. Do not go hunting for a fix on the dashboard.
MiningSettings › Mining Four tabs: Pools, Cooling, Performance, Advanced. Everything that decides how the machine runs lives behind this one menu item.Visited deliberately, not browsed.Three of these four tabs can cost you hardware. Advanced says so itself at the top of the page.
SystemSettings › System Four tabs: General, Security, Network, Maintenance. Housekeeping, plus the two buttons that undo everything.Password changed from the default, timezone correct, a backup taken.On 1.3.5 backup, restore, reset and rollback sit on Maintenance. The vendor's own System documentation still describes three tabs and puts them under General, so it will send you to the wrong place.
Tech supportTech support Opens a panel with four tiles: a diagnostic report, the API documentation the miner serves itself, a Telegram channel that is not live yet, and the vendor's documentation site.The diagnostic report is the one that earns its place.For a hardware fault this is the wrong door. A board that will not detect chips is not a firmware ticket - generate a diagnostic report - from here or from the Logs panel toolbar - and send it to us instead.
Firmware updateFirmware update Opens the flashing dialog over whatever page you were on.Opened when you actually intend to flash.Nothing happens until a file is dropped in and Flash miner is pressed - but that is the only safety net between here and a rewritten control board.
General tabSystem › General Timezone, interface language and appearance.Timezone matching where the miner physically is.Harmless except for the timezone, which quietly corrupts every log you ever send anyone.
Security tabSystem › Security Change password, and generate API keys for monitoring tools.Password changed from the default; one API key per tool so you can revoke one without breaking the rest.A miner on a shared or hosted network still using the default password is a miner someone else can repoint at their own pool.
Network tabSystem › Network Hostname, DHCP toggle, and manual IP, mask, gateway and two DNS entries. There is a Test network button.DHCP with a reservation on the router, or a static address you actually wrote down.Switch to static, get the gateway wrong, and the miner is unreachable - the only way back is physical. Press Test network before you walk away.
Maintenance tabSystem › Maintenance Backup, restore, reset to defaults and rollback to stock.A backup taken after every good tune.Two of the four buttons here throw away hours of tuning, and one of them takes the firmware off entirely.
TimezoneTimezone Sets the clock the logs are stamped with.Your actual local time zone.Leave it wrong and every log you send us is offset from the events you are describing. That costs diagnostic time, which on a repair is money.
Interface languageInterface language English, Russian, Ukrainian, Portuguese, Persian and Georgian.Whatever you read fastest.If you send us screenshots, English is easiest for us to work from.
ThemeTheme Light, dark, or follow the operating system.Whatever you can read.Nothing breaks here.
Dark sidebarDark sidebar Applies to the light theme only.Personal preference.Nothing breaks here either.
Optimized renderingOptimized rendering Disables animation and interface effects.On if the interface feels slow on your machine.Purely cosmetic - it changes nothing about how the miner runs.
Display chips number on boardDisplay chips number on board Numbers each chip position in the chip layout view.On, if you are diagnosing anything.Not a risk, just badly hidden. Turn it on before you screenshot a chip map for us - a numbered map tells us where on the physical board the problem sits.

Security

Two things live here, and both of them matter more than they look.

Vnish 1.3.5 System Security tab where the web interface password is changed and API keys are issued
DashboardMonitoring › Dashboard The monitoring screen. Nothing here changes anything - it is everything the miner is doing right now, on one page.Where you should be looking most of the time.It is read-only, which also means a problem visible here has to be fixed on one of the other screens. Do not go hunting for a fix on the dashboard.
MiningSettings › Mining Four tabs: Pools, Cooling, Performance, Advanced. Everything that decides how the machine runs lives behind this one menu item.Visited deliberately, not browsed.Three of these four tabs can cost you hardware. Advanced says so itself at the top of the page.
SystemSettings › System Four tabs: General, Security, Network, Maintenance. Housekeeping, plus the two buttons that undo everything.Password changed from the default, timezone correct, a backup taken.On 1.3.5 backup, restore, reset and rollback sit on Maintenance. The vendor's own System documentation still describes three tabs and puts them under General, so it will send you to the wrong place.
Tech supportTech support Opens a panel with four tiles: a diagnostic report, the API documentation the miner serves itself, a Telegram channel that is not live yet, and the vendor's documentation site.The diagnostic report is the one that earns its place.For a hardware fault this is the wrong door. A board that will not detect chips is not a firmware ticket - generate a diagnostic report - from here or from the Logs panel toolbar - and send it to us instead.
Firmware updateFirmware update Opens the flashing dialog over whatever page you were on.Opened when you actually intend to flash.Nothing happens until a file is dropped in and Flash miner is pressed - but that is the only safety net between here and a rewritten control board.
General tabSystem › General Timezone, interface language and appearance.Timezone matching where the miner physically is.Harmless except for the timezone, which quietly corrupts every log you ever send anyone.
Security tabSystem › Security Change password, and generate API keys for monitoring tools.Password changed from the default; one API key per tool so you can revoke one without breaking the rest.A miner on a shared or hosted network still using the default password is a miner someone else can repoint at their own pool.
Network tabSystem › Network Hostname, DHCP toggle, and manual IP, mask, gateway and two DNS entries. There is a Test network button.DHCP with a reservation on the router, or a static address you actually wrote down.Switch to static, get the gateway wrong, and the miner is unreachable - the only way back is physical. Press Test network before you walk away.
Maintenance tabSystem › Maintenance Backup, restore, reset to defaults and rollback to stock.A backup taken after every good tune.Two of the four buttons here throw away hours of tuning, and one of them takes the firmware off entirely.
Current passwordChange password › Current password The password the interface has now. Vnish ships with a well-known default.Something only you know.A miner on a hosted site or a shared network still on the default is a miner anyone on that network can repoint at their own pool. Nothing on the dashboard would tell you it happened.
New passwordChange password › New password What you are changing it to. Takes effect immediately.Changed once, written down somewhere you will still have in a year.Lose this and the interface is closed to you. Getting back in means a reset that takes the tuning with it.
API keysAPI keys › Add new key Generates a key so monitoring software can read the miner without being given the interface password.One key per tool, so revoking one does not break the others.A key you hand to a monitoring service is a key that service now holds. Issue them separately and revoke them when you stop using the tool - there is no expiry here.

Network

With DHCP on, the four fields below are greyed out. Turn it off and they become editable.

Vnish 1.3.5 System Network tab with hostname, DHCP toggle and the IP address, mask, gateway and DNS fields
DashboardMonitoring › Dashboard The monitoring screen. Nothing here changes anything - it is everything the miner is doing right now, on one page.Where you should be looking most of the time.It is read-only, which also means a problem visible here has to be fixed on one of the other screens. Do not go hunting for a fix on the dashboard.
MiningSettings › Mining Four tabs: Pools, Cooling, Performance, Advanced. Everything that decides how the machine runs lives behind this one menu item.Visited deliberately, not browsed.Three of these four tabs can cost you hardware. Advanced says so itself at the top of the page.
SystemSettings › System Four tabs: General, Security, Network, Maintenance. Housekeeping, plus the two buttons that undo everything.Password changed from the default, timezone correct, a backup taken.On 1.3.5 backup, restore, reset and rollback sit on Maintenance. The vendor's own System documentation still describes three tabs and puts them under General, so it will send you to the wrong place.
Tech supportTech support Opens a panel with four tiles: a diagnostic report, the API documentation the miner serves itself, a Telegram channel that is not live yet, and the vendor's documentation site.The diagnostic report is the one that earns its place.For a hardware fault this is the wrong door. A board that will not detect chips is not a firmware ticket - generate a diagnostic report - from here or from the Logs panel toolbar - and send it to us instead.
Firmware updateFirmware update Opens the flashing dialog over whatever page you were on.Opened when you actually intend to flash.Nothing happens until a file is dropped in and Flash miner is pressed - but that is the only safety net between here and a rewritten control board.
General tabSystem › General Timezone, interface language and appearance.Timezone matching where the miner physically is.Harmless except for the timezone, which quietly corrupts every log you ever send anyone.
Security tabSystem › Security Change password, and generate API keys for monitoring tools.Password changed from the default; one API key per tool so you can revoke one without breaking the rest.A miner on a shared or hosted network still using the default password is a miner someone else can repoint at their own pool.
Network tabSystem › Network Hostname, DHCP toggle, and manual IP, mask, gateway and two DNS entries. There is a Test network button.DHCP with a reservation on the router, or a static address you actually wrote down.Switch to static, get the gateway wrong, and the miner is unreachable - the only way back is physical. Press Test network before you walk away.
Maintenance tabSystem › Maintenance Backup, restore, reset to defaults and rollback to stock.A backup taken after every good tune.Two of the four buttons here throw away hours of tuning, and one of them takes the firmware off entirely.
HostnameHostname The name shown in the browser tab and in network scans. Ours reads Antminer.Something that identifies this specific machine.Every miner on the site called Antminer means the browser tab tells you nothing about which one you are looking at, and neither does your scanner.
DHCPDHCP - Automatic network setup Let the router hand out the address. On by default, and while it is on the manual fields are locked.On, with a reservation on the router so the address stops moving.The address can change after a power cut. If you have bookmarked an IP, it is the bookmark that breaks, not the miner.
IP Address and MaskIP Address / Mask The manual configuration, active only when DHCP is off.Filled in from what the network actually uses.An address outside the subnet makes the miner unreachable the moment you save. From here there is no undo button - only physical access.
DNS 1, DNS 2 and GatewayDNS 1 / DNS 2 / Gateway Where the miner sends name lookups and everything not on the local network.A gateway that exists, and two DNS servers rather than one.A wrong gateway is the classic way to lose a miner. The pool goes unreachable, the interface stays up on the local network, and it looks like a pool problem for the first hour.
Test networkTest network Runs a connectivity check from the miner itself.Pressed before you walk away from any change on this tab.The one control here that costs nothing and saves a trip to wherever the machine lives.

Maintenance

Backup, restore, reset and rollback to stock all live on this tab. The vendor's own documentation still describes a three-tab System section with these under General Settings.

Vnish 1.3.5 System Maintenance tab with backup settings, restore, reset and rollback to stock firmware
DashboardMonitoring › Dashboard The monitoring screen. Nothing here changes anything - it is everything the miner is doing right now, on one page.Where you should be looking most of the time.It is read-only, which also means a problem visible here has to be fixed on one of the other screens. Do not go hunting for a fix on the dashboard.
MiningSettings › Mining Four tabs: Pools, Cooling, Performance, Advanced. Everything that decides how the machine runs lives behind this one menu item.Visited deliberately, not browsed.Three of these four tabs can cost you hardware. Advanced says so itself at the top of the page.
SystemSettings › System Four tabs: General, Security, Network, Maintenance. Housekeeping, plus the two buttons that undo everything.Password changed from the default, timezone correct, a backup taken.On 1.3.5 backup, restore, reset and rollback sit on Maintenance. The vendor's own System documentation still describes three tabs and puts them under General, so it will send you to the wrong place.
Tech supportTech support Opens a panel with four tiles: a diagnostic report, the API documentation the miner serves itself, a Telegram channel that is not live yet, and the vendor's documentation site.The diagnostic report is the one that earns its place.For a hardware fault this is the wrong door. A board that will not detect chips is not a firmware ticket - generate a diagnostic report - from here or from the Logs panel toolbar - and send it to us instead.
Firmware updateFirmware update Opens the flashing dialog over whatever page you were on.Opened when you actually intend to flash.Nothing happens until a file is dropped in and Flash miner is pressed - but that is the only safety net between here and a rewritten control board.
General tabSystem › General Timezone, interface language and appearance.Timezone matching where the miner physically is.Harmless except for the timezone, which quietly corrupts every log you ever send anyone.
Security tabSystem › Security Change password, and generate API keys for monitoring tools.Password changed from the default; one API key per tool so you can revoke one without breaking the rest.A miner on a shared or hosted network still using the default password is a miner someone else can repoint at their own pool.
Network tabSystem › Network Hostname, DHCP toggle, and manual IP, mask, gateway and two DNS entries. There is a Test network button.DHCP with a reservation on the router, or a static address you actually wrote down.Switch to static, get the gateway wrong, and the miner is unreachable - the only way back is physical. Press Test network before you walk away.
Maintenance tabSystem › Maintenance Backup, restore, reset to defaults and rollback to stock.A backup taken after every good tune.Two of the four buttons here throw away hours of tuning, and one of them takes the firmware off entirely.
Backup settingsCreate archive Downloads your configuration as an archive.Taken once after a good tune, and again before you change anything significant.There is no cost to having one, and a three-to-four hour retune if you needed one and did not take it.
Restore SettingsRestore Loads a configuration archive you saved earlier.An archive from the same model and the same firmware branch.Restoring a backup from a different model or branch pushes settings the hardware was never tuned for.
ResetFirmware Reset › Reset Returns the firmware to its own defaults. You stay on Vnish.A reasonable first move when a machine behaves oddly and you have lost track of what was changed.Your tune goes with it. Take a backup first and plan for a retune.
Rollback to stockFirmware Reset › Rollback to stock Puts Bitmain's original firmware back.Only when you actually want to be on stock.Going back is not always as easy as coming forward - on recent Amlogic boards with security firmware the return path can be closed. And a rollback does not restore the warranty; Bitmain's position is that it was voided when the third-party firmware went on.

Tech support

Opens as a panel over whatever page you were on. Four tiles: one that matters, one that documents the machine's own API, one placeholder and one link out.

Vnish 1.3.5 tech support dialog on an Antminer S21 Pro with the diagnostic report, API documentation, Telegram channel and home page tiles
Diagnostic reportDiagnostic report Packs the firmware's own view of the machine into one archive: logs, the hardware scan, chain and pool state, and the current settings.Generated while the fault is happening and attached to your message.Take it before you reboot. A report pulled after a restart describes a healthy miner, and the chain that was failing an hour ago is not in it. The same button sits in the Logs panel toolbar.
APIAPI Opens the API documentation the miner serves itself, at /docs on its own address.Every screen in this guide is also reachable over HTTP, which is how a room full of miners gets managed without clicking through each one.It documents the build sitting in front of you rather than the newest release, so it is the version to trust when the two disagree.
Telegram channelTelegram channel - Coming soon Marked Coming soon in 1.3.5. The tile is greyed out and does nothing.Nothing. It is a placeholder.Noted here only so you do not spend time working out why it will not open.
Home pageHome page Opens the firmware vendor's documentation site in a new tab.The right place for questions about the firmware itself - what a setting does, what changed in a release.A hardware fault is not a firmware ticket. A board that will not detect its chips, a chain that drops under load, a PSU that will not hold voltage - those need the diagnostic report from the tile above, sent to whoever repairs the machine.
Version and patent line2026 - Vnish (c) / pat. us-11631138 The footer of the panel: the vendor's own branding and the patent the firmware ships under.A quick confirmation that what is installed is Vnish rather than stock or another custom build.Worth a glance on a machine you did not flash yourself, before you start explaining behaviour by the wrong firmware.

Firmware update

Opens as a dialog over whatever page you were on. The header badges restate the model, platform and current version.

Vnish 1.3.5 firmware update dialog on an Antminer S21 Pro with the firmware file selector, the configuration backup option and the flash miner button
Header badgesFirmware update The dialog restates what it thinks the miner is: model, control board platform, current firmware.A platform badge matching the file you are about to upload.This is your last free check before flashing. If the badge says aml and your file is a Xilinx build, close the dialog.
Save current miner configurationSave current miner configuration Keeps your current settings across the flash instead of coming back to defaults.On, which is the default.Turn it off deliberately if you are flashing precisely because the configuration is suspect - carrying a bad config across a reflash defeats the point.
Firmware fileDrop file here or click to upload Where the build goes. The file name appears underneath once it is loaded.A file whose name matches your model, your board family and the version you meant to install.Read the file name back before you continue. Almost every wrong-file incident is visible right here and gets clicked past.
Flash minerFlash miner Starts writing. Disabled until a file is chosen.Pressed once, on a machine with stable power.Do not cut power, unplug anything or close the lid from here until it finishes. An interrupted write leaves the control board without working low-level software, and that is a service-centre job rather than a retry.

Antminer Repair LLC — USA-based ASIC repair lab, Fort Lauderdale, Florida. Official Vnish partner and authorized distributor of custom firmware for Bitmain Antminer.
Screens captured from Vnish 1.3.5, build VnishnetMRepair, compiled 2026-07-31, on an Antminer S21 Pro with an Amlogic control board and a NAND install. Other models differ in the ranges, in which boards are present, and in which settings exist at all: some controls appear only on miners with a PIC controller, others only on miners without one.
Firmware downloads · Remote install service · Hashboard repair, USA · Contact