DiagnosticsIntermediate11 min read

Miner Health Check: Is Your ASIC Performing at 100%? (2026)

Published July 24, 2026

A miner that is connected is not necessarily a miner that is healthy. It can be quietly running 20% down on a throttled hashboard, or throwing rejects into a slow connection, and still show up on the dashboard. This guide is a routine you can run in a few minutes to confirm your ASIC is actually pulling its full weight, checking both the miner itself and what the pool sees.

Start at the pool

The fastest health signal is your worker on the mkpool dashboard, because it is measured independently of the miner's own reporting. Open your address on the miners page and check three things, all explained in full in reading your worker stats:

Hashrate, the 1-hour or 1-day average near the miner's rating. Ignore the current figure for this; it is too jumpy to judge health.
Reject rate, invalid and stale shares as a small fraction of valid ones.
Last share, only seconds old, confirming the worker is live right now.

If all three look right, the miner is doing its job and the rest is variance. If one is off, the split below tells you where to look.

Then check the miner

Open your miner's own dashboard (AxeOS, the Antminer or Whatsminer web UI, Braiins OS, and so on) and work down this list:

Hashrate vs rating. After a 10-15 minute warm-up, the miner's average should sit near its rated figure. A steady shortfall is the headline symptom the rest of this list explains.
Every hashboard reporting. Multi-board ASICs show each board separately. One board reading zero removes its whole share of the total, a third of the hashrate on a three-board miner, even though the miner still "works".
Temperatures in the safe range. Chips throttle themselves when too hot, which shows up as a lower hashrate. Check your model's rated safe range; sustained operation near the top of it, or a rising trend, means cooling is the limit.
Fans spinning normally. A fan stuck at maximum is fighting heat; a fan reading zero or erroring risks a thermal shutdown.
Power stable. Draw close to the expected figure and no voltage warnings. An undersized or failing supply sags under load, which destabilises hashrate and triggers resets.
Power mode / tuning. Confirm the firmware is not in a low-power or eco profile that caps hashrate by design.

Symptom → likely cause → fix

What you see	Most likely cause	What to do
Hashrate a few % low	Warm-up, normal variance, minor throttling	Recheck after 15 min; improve airflow
Hashrate well below rating	Thermal throttle, weak PSU, or dead board	Check temps and fans, then power, then per-board readings
One board reads zero	Loose cable, overheating, board fault	Power down, reseat data/power connectors; RMA if still dead
High invalid-share rate	Difficulty misconfig or failing chip	Use the vardiff port; inspect the board if it persists
High stale-share rate	Network latency	Use a nearer endpoint; check for packet loss or Wi-Fi drops
Frequent disconnects / reboots	Power brownout or thermal protection	Resize the PSU; improve cooling; check the network link
Worker offline on the pool	Power, network, or a crashed miner	Confirm power and network, then power-cycle the miner

Don't forget the room

Half of all "my miner is slow" problems are environmental. Heat the miner cannot get rid of is heat that throttles it, so before blaming the hardware, check the basics: enough clear space around intake and exhaust, a reasonable ambient temperature, no recirculating hot air, and clean dust filters. A miner that throttles in a hot cupboard will run at full speed in a ventilated room. Cooling options are compared in the cooling guide.

A clean bill of health

1-hour and 1-day hashrate near the rated figure.
Every hashboard contributing.
Temperatures in range, fans not maxed.
Reject rate near zero; last share seconds old.
Power draw as expected, no voltage warnings.

When not to worry

A healthy miner that has found no blocks is not a broken miner. Shares prove it is working; blocks are the lottery. If your health check passes, long gaps between blocks are just variance, not a fault, and you can put real numbers on the wait with the chance calculator.

Disclaimer

Safe temperature ranges, dashboard fields and firmware controls vary by miner and firmware version; follow your device's documentation for exact figures. Details are current as of July 2026. Nothing here is financial advice. Solo mining is probabilistic and can lose money depending on electricity cost and network conditions.

Frequently asked questions

How do I know if my miner is running at full hashrate?

Let it warm up for 10 to 15 minutes, then compare the miner's own average hashrate against its rated figure, and cross-check the pool's 1-hour average for your worker. A small shortfall is normal; a persistent gap of more than a few percent usually means thermal throttling, an underpowered supply, or a hashboard that has dropped out.

Why is my hashrate lower than the rated number?

The three usual causes, in order: thermal throttling (the chips slow themselves when too hot), an undersized or failing power supply (voltage sags under load), and a dead or partially failed hashboard (which removes a fixed fraction of the total). Check temperatures and fans first, then power, then whether every hashboard is reporting.

What reject rate is normal?

Invalid and stale shares should be a small fraction of valid shares, low single-digit percent or less. A few are always expected. A climbing reject rate is the signal to investigate: a high invalid rate suggests a difficulty misconfiguration or failing hardware, while a high stale rate suggests network latency.

My miner keeps disconnecting or restarting. What causes that?

Repeated disconnects or reboots usually trace to power or heat: a supply that browns out under load, or thermal protection cutting in. Network problems and a failing hashboard can also cause it. Check that the supply is rated well above the miner's draw, that airflow and ambient temperature are adequate, and that the network link is stable.

One hashboard shows zero. Is the miner broken?

A hashboard reading zero means that board is not contributing, so you lose its share of the total hashrate. It can be a loose cable, overheating, or a hardware fault. Power down, reseat the board's data and power connectors, and check temperatures; if it stays dead, the board likely needs repair or replacement under warranty.

Newsletter

Mining insights in your inbox

New guides, hardware reviews, and pool updates. No spam, unsubscribe anytime.

Keep reading

DiagnosticsBeginner11 min read

Reading Your Worker Stats on mkpool (2026)

What every number on your mkpool dashboard means: the hashrate windows, valid/invalid/stale shares, current difficulty, last share, and best share, including why best share is a record of how close you came, not progress toward a block.

Mecanik1337Read →

Setup & First ShareIntermediate12 min read

Antminer S21 Series: Solo Mining Setup on mkpool (2026)

Point an Antminer S21, S21 Pro, S21 XP or S21 Hydro at mkpool and solo mine straight to your own wallet. Web UI access, exact stratum ports, difficulty tiers, TLS and Stratum V2, AsicBoost, fleet naming and how to verify the connection.

Mecanik1337Read →

Setup & First ShareBeginner11 min read

Connect Any Miner to mkpool: The Universal Setup Guide (2026)

One pattern connects every miner to mkpool: a stratum URL, your own wallet address as the username, and the password x. Match your hardware's algorithm to a chain, pick a port, and mine straight to your wallet, no account, no custody.

Mecanik1337Read →

← back to all guides