The most common failures – and how long miners really last

Anyone buying a miner wants to know how long it will run. The short answer: a very long time, if it survives the first few weeks.

Early or not at all

Manufacturing defects show up quickly under full load. A bad solder joint or a weak chip fails within the first days to weeks. Whatever survives this phase lasts an extremely long time with a clean power supply and adequate cooling. Units that have been running non-stop for many years are not unusual.

For you that means: run a new unit under full load straight away and watch it closely in the first few weeks. During this time the manufacturer warranty applies.

What fails

Hashboards. The most common and most expensive case. The chips on a board are connected in series; if one fails or a solder joint loses contact, the whole board stops. In the log you can tell by a chip count that is too low or by missing temperature readings.

Fans. Wear parts with bearings. They announce themselves with noise and are replaced in minutes. A stopped fan must be replaced immediately, because otherwise the boards overheat.

Power supplies. Sensitive to unstable voltage, dust and high intake air temperature. Replacement is simple, but damage caused by wrong voltage is not a warranty case.

Control board. Rare. Usually it affects the storage, and reflashing the firmware fixes it.

Continuous operation and solar operation

Every time it is switched on, the hashboard heats up; every time it is switched off, it cools down. Chip, solder and circuit board expand by different amounts in the process. Each cycle puts a little stress on the solder joints.

A unit in continuous operation stays at one temperature and almost never experiences this stress. A unit that runs on solar power only during the day goes through it once every day.

The S17 series

With the Antminer S17 and T17 series this was a real problem. The heatsinks were glued to the chips, and this bond did not hold. Heatsinks fell off, chips overheated, loose parts caused short circuits. Samson Mow of Blockstream reported in 2020 that customers were reporting failure rates of 20 to 30% instead of the usual 5% or so. There is no independent series of measurements on this.

Today

In the generations that followed, failures caused by temperature changes hardly occur any more. Daily solar operation is no longer a significant risk with current units. If you still want to keep the cycles small, throttle the unit via firmware when there is little sun instead of switching it off completely. The temperature then stays more even.

What really determines the service life

1. Stable voltage. Fluctuations and overvoltage destroy power supplies and boards.
2. Clean, dry intake air. Dust insulates the heatsinks, moisture corrodes the contacts.
3. Temperature. Every degree less at the chip extends the life of the solder joints.
4. Moderate overclocking. More voltage means more heat and more stress.

Sources

• Cointelegraph: failure rates of the Antminer S17 and T17, April 2020
• Deng et al.: failure of solder joints under thermomechanical load, Materials 2026
• D-Central: thermal fatigue on hashboards

As of October 2026.