Decoding Antminer (Bitmain) logs

A complete guide to decoding Antminer (Bitmain) logs

Bitmain Antminer Logs

Log	What it means	What to do
Chain[X]: find 0 asic or chain find 0 asic, will power off this chain	The board is not visible at all / no chips detected	1. Power off the ASIC. 2. Pull out and firmly reseat both thick power cables on this board. 3. Power on. In 40–60% of cases the S19kPro/XP/S21 immediately finds 126 chips and starts working. Doesn’t help → send the board for repair.
Chain[X]: only find 47–90 asic	Half of the chips are missing	Reseat the power cables and the ribbon cable. Doesn’t help → send the board for repair.
autotune fail (especially on S21/S21+)	The board could not tune its frequencies	1. Reseat the cables. 2. Factory Reset + fresh firmware. 3. Doesn’t help → in 85% of cases the VRM (the power block on the board) is dead → send for repair.
domain voltage abnormal / unstable	Voltage on the board has sagged by about 1.2 V	Send the board for repair.
CRC error (even 1–2 times a minute)	Chips are passing data to each other poorly	This is the beginning of the end. Within 1–4 weeks the board will drop to zero. Send it for repair.
nonce error	A chip is computing “garbage”	On older S9–S17 — a dead chip. On S19kPro and newer, in 90% of cases it’s bad power to the board.
Temp[X]: 90–97 °C on S21	Normal for the S21 in 2025	Do nothing — this is the normal state.
Temp[X]: 95–100+ °C (any model)	About to shut down	Urgently improve airflow and remove dust.
Fatal Error: some fan lost!	One fan is not spinning or not reporting speed	Swap the fans around → if the error moved, replace the fan. If the error stayed on the same number → controller repair.
PSU error / voltage too low	The PSU is not working or there is a voltage drop at the outlet	Remove splitters, check for 220 V, try a different PSU.
pic error / pic init fail	The small “brain” (PIC) on the board has burned out	Send the board for repair.
EEPROM error / eeprom checksum fail	The board’s memory is dead or the data has been wiped	In 90% of cases the board needs repair. Reflashing almost never helps.

Miner Startup and Software

Log line	What it means	What to do
bmminer start / cgminer start	The main miner program has started	Normal
restart bmminer	The miner was restarted inside the system (without a full ASIC reboot)	Look at the lines above — find out why it restarted
Found bmminer / start running bmminer	The controller found and launched the miner binary	Normal
bmminer not running, restart…	The process crashed, the watchdog is trying to bring it back up	Look higher up in the log (temperature, chips, power)

Board Detection (Chain / Hashboard)

Line	What it means	What to do
check_chain_num: chain_num = 3	The ASIC expects 3 hashboards (0, 1, 2)	Normal
chain[0] exist / chain[1] exist	The board was found via its cable	Normal
chain[2] is missing / doesn’t exist	One of the boards is not responding at all	Check power, cables, port. Most likely — repair.
hashboard type: S19	The controller identified the board type/model	Normal
hashboard missing	The board is absent or “dead”	Check cables and power. Doesn’t help — send for repair.

ASIC (Chip) Detection

Line	What it means	What to do
chain[0]: find 126 asic	126 chips found on board 0 — normal for the S19 series	All good
chain[1]: find 0 asic	The board is physically there, but no chip responds	Check power and power cables (on S19kPro/S19XP/S21 a poor contact in the connector is common). If that doesn’t help — repair.
chain[1]: only find 47 asic	The board found only some of the chips; the chain breaks off	Check the ribbon cable. If it persists — chain/chip repair.
chain[2]: asic detected = 76	Alternative format: 76 chips found	If it should be 76 — normal. If more — a faulty chip, repair.
asic number is not enough	Fewer chips found than expected	The board is faulty — diagnostics needed.
chain 1 find asic 0, will power off this chain	The controller is disabling this board because no chips were found	The ASIC will run without this board (hashrate will drop). Repair the board.

PIC, Sensor, and Service Controller Initialization

Line	What it means	What to do
pic init / init pic ok	The PIC controller started successfully	Normal
pic error / failed to init pic	The power/sensor controller is not communicating	Most often — board repair. Reflashing sometimes helps, but not always.
read temp sensor ok	The temperature sensors are responding	Normal
read temp sensor error / temp sensor read fail	Unable to read temperature from the board	Check the ribbon cable. If the error is consistently on one board — repair.

Fan Check at Startup

Line	What it means	What to do
check fan NUM: 2	2 fans expected (or 4 — depends on the model)	Normal
Fan0 speed = 0, Fan1 speed = 0 (startup phase)	The fans haven’t spun up yet at startup	Normal, as long as normal values appear later
Fatal Error: some fan lost! (already at startup)	The controller immediately sees that one fan is not responding	Check the fan, cable, connector. Without a fix the ASIC will not start mining.

Calibration and Autotune at Startup

Line	What it means	What to do
start autotune / begin tuning chain 0	Selection of frequencies and voltages for the boards begins	Normal
tune chain 0 asic 0-125	All chips on the board are being tuned	Normal
autotune fail (especially on S21)	The board could not tune its frequencies	Check cables and reflash. Doesn’t help → VRM problem → send for repair.
calibration OK	Calibration completed successfully	Normal
calibration fail	The board could not be calibrated	Restart. If it repeats — repair.

What You See in the Logs — Decoding

What you see in the logs	What it means	What to do
chain 1 lower hash	This board is performing worse than the others, giving fewer terahashes	1. Compare this board’s temperature with the others. 2. Check how many chips it found at startup. 3. Are there CRC errors? If so — the board is degrading.
only find 47 asic	The ASIC found fewer chips. The board is incomplete / partially dead.	The board works, but is already broken. Needs diagnostics or repair.
CRC error on chain 1	Chips on the board are transmitting data with errors. The chain is unstable.	If CRC errors repeat — the board is dying. You can mine temporarily, but the hashrate will keep dropping → repair.
disable chain 1	The ASIC disabled the board itself — it was in very bad shape	Nothing can be done. The ASIC will mine only on the remaining 1–2 boards. Repair the board.
asic X timeout / nonce error	One of the chips is glitching, responding poorly, or computing incorrectly	An occasional error is normal. Frequent ones → the board is degrading → repair.
hashrate is now much lower than avg	The hashrate is jumping, the ASIC is unstable	Check: 1) temperature; 2) whether one board is performing worse; 3) whether there is CRC; 4) whether the ASIC is throttling due to heat.

Common Power Errors

What you see in the logs	What it means	What to do
voltage too low	The ASIC isn’t getting enough power. Voltage has sagged.	1. Measure the voltage at the outlet (should be 220–240 V). 2. Remove extension cords and splitters. 3. Check the power cable. 4. If the problem remains — replace the PSU.
voltage too high	Voltage above normal is dangerous for the ASIC	Check the outlet with a multimeter. If the PSU is reporting this — replace the PSU.
PSU error	The power supply is malfunctioning	1. Check the cable and contacts. 2. Replace the PSU — this is a clear sign it is dying.
power lost	Power briefly disappeared / failed	1. Check the outlet and cable. 2. Check the breaker / panel. 3. If everything is fine — replace the PSU.
power low, reboot	The ASIC is rebooting due to a power sag	Check the PSU, replace the PSU.

Internal Board Power (Domain Voltage)

What you see	What it means	What to do
domain voltage abnormal	Voltage on one of the boards is jumping or unstable. The problem is in the board itself, not the outlet.	The board needs repair. It can’t be fixed by yourself.
set volt xx / volt adjust	The ASIC is selecting voltage for the chips — normal	Nothing needs to be done.
volt too low for chain X	This board isn’t getting enough power to run stably	Check the PSU. If the PSU is fine — the board is degrading, repair needed.

Power Limit

What you see	What it means	What to do
power limit reached	The ASIC reduced its power on its own because the PSU can’t keep up or the grid is sagging	Check: 1) PSU temperature; 2) 220 V mains; 3) if the problem remains — replace the PSU.
limited power mode	Economy mode is active — the hashrate will drop	Check the power supply.

Pool Connection

What you see	What it means	What to do
stratum connect	The ASIC is trying to connect to the pool	Normal
stratum connected	Connection to the pool succeeded	All good — it’s mining.
stratum authenticate success	The ASIC is authorized (Worker is correct)	Normal

Network Issues (Cable, Internet, DNS)

What you see	What it means	What to do
socket connect failed	The ASIC can’t connect to the pool. No internet.	1. Check the cable. 2. Check the router. 3. Try a different pool.
reconnecting…	The ASIC is losing its connection and trying to reconnect	The internet is unstable. Check the network.
DNS resolve failed	The ASIC can’t find the pool’s domain (DNS problem)	Change the DNS on the router to 8.8.8.8 / 1.1.1.1
stratum connection interrupted	The connection was lost	Check the internet, cable, connector.
network error, retry after 10 seconds	There is internet, but it’s unstable	Check ping/latency.

Pool Problems (Authorization or the Pool Itself Is Down)

What you see	What it means	What to do
stratum authentication failed	Wrong Worker or password (usually “x”)	Check the worker string on the pool.
pool 0 slow/down	The pool is not responding or is slow	Switch pools / wait.
null work received	The pool sent an empty job	Problem on the pool’s side.
pool failed: retry after …	The ASIC can’t exchange data with the pool	Check the network → try a different pool.

Share and Acknowledgment Errors

What you see	What it means	What to do
accepted share	The pool accepted the share — everything works	Normal
rejected share	The pool is rejecting some shares	If >3% → check the internet (latency).
stale share	Outdated shares — the ASIC submitted them too late	Network latency problem. Check your ISP.
high stale rate	Too many stale shares	Internet speed is insufficient. Change the router or cable.

Key Symptoms When the ASIC “Works” but Isn’t Mining

What you see	What it means	What to do
hashrate now = 0, but fans are spinning	The ASIC isn’t receiving jobs from the pool	Check the network / pool / DNS.
No accepted share in 1–2 minutes	The ASIC is connected but isn’t doing any work	Pool or authorization problem.
Constant reconnecting	The internet is “floating”	Check the cable, router, change the port.
socket send error	The ASIC can’t send shares to the pool	Check the network.

Most Common Causes of Reboots (Watchdog)

What you see in the logs	What it means	What to do
watchdog: restart miner	The ASIC froze or stopped hashing → the watchdog forced a restart	Look 10–30 lines above: temperature? CRC? power? network?
restart bmminer	The miner itself crashed (as a program)	Often due to bad power or chip errors. Check the PSU → check CRC.
power lost, reboot	Power disappeared/sagged, the ASIC restarted	Check the outlet, cable, PSU.
fatal error, rebooting	A hard hardware error	See which chain or sensor failed → usually repair.

Reboot Due to Temperature

What you see	What it means	What to do
temp too high, reboot	The ASIC overheated and the protection kicked in	Check ventilation, dust, room temperature.
overheat, shutdown, reboot	The chips or board exceeded the limit	Cool it down first, then test.
fan lost, reboot	A fan isn’t working → protection → reboot	Check the fans, cables, ports.

Reboot Due to Power (PSU, 220V Sag)

Log	What it means	What to do
power low, reboot	The PSU can’t cope, low voltage	Check the 220V mains, remove extension cords, replace the PSU.
voltage too low, restart	Severe power sag	Same as above.
psu error, reboot	The power supply is malfunctioning	In 95% of cases → replace the PSU.

Reboot Due to Chips / Board (Chain)

What you see	What it means	What to do
chain X asic status abnormal, reboot	The board froze or the chips aren’t responding	Look at CRC, temperature, and the problem chip.
CRC error … restart	The chain is falling apart, chips are transmitting errors	Board repair.
chain X lower hash, restart	The board’s hashrate dropped → the watchdog treats it as a failure	Check the board (temperature, CRC, bad chip).
fault on chain X, reboot	Hard chain error	Send the board for repair.

Reboot Due to Network / Pool

What you see	What it means	What to do
stratum disconnect, reboot	The ASIC loses the pool and considers it “dead”	Check the internet, DNS, router.
no active pool, reboot	The ASIC didn’t find a single working pool	Fix the worker string, change the pool.

PIC Errors (Microcontroller on the Board)

What you see in the logs	What it means	What to do
read pic fail	The controller isn’t receiving data from the PIC	Send the board for repair. Can’t be fixed at home.
pic init fail	The PIC couldn’t start at boot	Board repair / PIC reflash on a test bench.
pic sensor error	The PIC can’t read the temperature sensor	Check the ribbon cable → if that doesn’t help, board repair.

Temperature Sensor Errors

What you see	What it means	What to do
temp sensor error	The ASIC can’t see the temperature sensor	Check the cables. If they’re connected to the board — repair.
read temp fail	Sensor read error	Same as above.
sensor connection lost	The sensor isn’t responding	Check the cable → board repair.
board temp abnormal	The sensor is giving incorrect readings	The board is unstable → send for repair.

Controller Errors (Control Board)

Log	What it means	What to do
control board lost chain	The controller doesn’t see the board	Reseat the cables. Doesn’t help — board or controller repair.
control board reboot	Controller failure	Reflash / Factory Reset.
control board error	The controller is unstable	Reflash. If it persists — replace the controller.
gpio init error	The controller can’t initialize the board control lines	Typical for S17/T17. Almost always a board repair.

EEPROM Errors

In short: if you see the word EEPROM in the logs — send the board for repair right away.

What you see	What it means	What to do
EEPROM error	The EEPROM can’t be read or is damaged	Board repair or EEPROM reflash with a programmer.
eeprom data invalid	The data is corrupted	Same as above.
eeprom not found	The EEPROM isn’t detected	The board is faulty → repair.

Model-Specific Logs

ANTMINER S19, S19 Pro, S19j, XP

Model	Specific error	What it means	What to do
S19 series	start autotune	Frequency autotune launched	Normal
S19 series	start voltage tune	Chain voltage tuning	Normal
S19 series	voltage calibration	Voltage calibration on the boards	Monitor it, as long as the error doesn’t repeat.
S19/S19j	asic index abnormal	Chip order failure	Check CRC, possibly repair.
S19/S19j	hashboard mismatch	The board doesn’t match the firmware/model	Check the firmware and compatibility; sometimes the board isn’t original.
S19/S19 XP	domain voltage abnormal	Bad chain power	Send the board for repair.
S19 Pro	pic init fail	The PIC doesn’t start	Board repair.

ANTMINER S21, T21, PRO

Model	Specific error	What it means	What to do
S21/T21	VRM protection triggered	VRM power protection triggered	The board is faulty → repair.
S21/T21	overcurrent on chain	Too much current on the board	Check the PSU, most likely repair is needed.
S21/T21	domain voltage unstable	Voltage is jumping	The board is degrading → repair.
S21/T21	chip detect timeout	Chips didn’t initialize in time	Check the cable → board repair.
S21/T21	critical temp: hashing stopped	Temperature reached a critical level	Check the cooling.

ANTMINER L7 (LTC/DOGE)

Model	Specific error	What it means	What to do
L7	L7: pll error	The PLL frequency can’t be set	Board repair or lower the frequency.
L7	L7 autotune fail	Autotune couldn’t find suitable frequencies	Check power and temperature.
L7	chip temp abnormal	Problem with the board’s temperature sensor	Check the sensors → repair.
L7	fan lost	Losing a fan is critical for the L7	Check the fans.

ANTMINER HYDRO

Model	Specific error	What it means	What to do
S19 Hydro / S21 Hydro	water temp high	The water is overheated	Check the cooling loop.
Hydro	pump speed low	The pump is running too slowly	Check the pump.
Hydro	flow sensor error	No water flow	Stop operation → loop repair.
Hydro	water cooling protect	Water cooling protection triggered	Fix the cooling problem.
Hydro	ERROR_WATER_REVERSE	Protection triggered (lines are swapped)	Swap the supply and return lines.