Definition
A hashboard (also called a hash board, hashing board, or mining board) is the printed circuit board that does the actual proof-of-work inside an ASIC miner. It carries the mining chips, their power-delivery circuitry, temperature sensors, and the signal interface back to the control board. A typical Antminer carries three identical hashboards wired in parallel; the machine’s total hashrate is the sum of all three.
Each hashboard packs dozens to over a hundred ASIC chips into a single daisy-chained string. Commands flow into the chain on the TX/CI line, responses return on the RX/RO line, and a shared clock plus reset signal keep the whole string in lockstep. That entire conversation between the board and the brain of the miner travels over an 18-pin flat ribbon cable connecting each board to the control board, which runs Linux and the mining daemon (cgminer/bmminer) that schedules work.
How power reaches the chips
Hashboards don’t power each chip individually. ASIC chips run on extremely low core voltages, so the chips are grouped into series-wired voltage domains. The board takes the 12-15V coming from the PSU, runs it through a boost circuit, and then steps it down through LDOs so each domain receives a fraction of the total. The exact layout changes with every chip generation:
- Antminer S19 (BM1398): 76 chips across 38 domains (2 chips per domain), each domain around 0.36V, fed from a 14V-to-19V boost.
- Antminer S19 Pro (BM1398): 114 chips across 38 domains (3 chips per domain), each domain near 0.32V, with a 12.6V-to-20V boost.
- Antminer S21 / T21 (BM1368): 108 chips across 12 domains (9 chips per domain) at roughly 1.2V, boosted toward ~25V, with no PIC chip on the board.
- Antminer S21 XP (BM1370): 91 chips across 13 domains (7 chips per domain) near 1.04V, boosted to ~21V, using 12 level shifters.
This series-domain design is why a hashboard is so unforgiving. Because the chips in a domain share current, one failed chip or one bad regulator can drop the whole domain offline, and a broken link anywhere in the daisy chain can silence every chip downstream of it. That is the difference between a board reading “0 ASIC” and one running at a fraction of its rated hashrate.
Why hashboards fail (and how they get fixed)
Hashboard faults are the single most common cause of lost hashrate on an ASIC miner. The usual suspects are dead or short-circuited chips, burned voltage regulators, cracked solder joints (often from thermal cycling or rough shipping), heatsinks that have fallen off and left chips to overheat, and water or corrosion damage on hydro and immersion units. Because the chips are daisy-chained, a single open circuit in the middle of the string can take out an entire board’s reported chip count even when most of the silicon is still good.
Diagnosing this is part detective work, part electronics. Technicians read the chip enumeration the control board reports, then probe the board with a multimeter in diode mode against known-good reference values, hunting for the exact domain or chip where voltage drops out. Repair from there is fine-pitch surgery: hot-air rework to lift and replace a dead chip, reflowing cold solder joints, swapping a failed regulator, or replacing a damaged temperature sensor. It is precise, microscope-and-tweezers work, which is exactly why hashboard repair is treated as its own specialist discipline rather than a casual home fix.
Understanding the hashboard matters because it directly drives uptime and profitability. A miner advertised at a given terahash figure only hits that number when all three boards are fully populated and every domain is holding voltage. When one board drops, you lose roughly a third of your hashrate. Whether you are running a single Bitaxe on the bench or a rack of S21s, knowing how the board is structured tells you what a fault report actually means and whether a machine is worth repairing or retiring.
D-Central has been repairing Antminer hashboards since the early ASIC era, and that bench experience feeds straight back into the firmware and tooling we build. If you have a board reading low or showing dead chips, our team can diagnose and repair it. To learn the surrounding hardware, read up on the control board that coordinates the boards, the ASIC technology behind the chips, and the PCB fundamentals every hashboard is built on. Ready to keep your fleet hashing? Start with our ASIC troubleshooting guides.
In Simple Terms
The main mining board inside an ASIC containing rows of mining chips. Most miners have 3-4 hashboards.
A hashboard (also called a hash board, hashing board, or mining board) is the printed circuit board that does the actual proof-of-work inside an ASIC miner. It carries the mining chips, their power-delivery circuitry, temperature sensors, and the signal interface back to the control board. A typical Antminer carries three identical hashboards wired in parallel; the machine's total hashrate is the sum of all three.
Each hashboard packs dozens to over a hundred ASIC chips into a single daisy-chained string. Commands flow into the chain on the TX/CI line, responses return on the RX/RO line, and a shared clock plus reset signal keep the whole string in lockstep. That entire conversation between the board and the brain of the miner travels over an 18-pin flat ribbon cable connecting each board to the control board, which runs Linux and the mining daemon (cgminer/bmminer) that schedules work.
How power reaches the chips
Hashboards don't power each chip individually. ASIC chips run on extremely low core voltages, so the chips are grouped into series-wired voltage domains. The board takes the 12-15V coming from the PSU, runs it through a boost circuit, and then steps it down through LDOs so each domain receives a fraction of the total. The exact layout changes with every chip generation:
- Antminer S19 (BM1398): 76 chips across 38 domains (2 chips per domain), each domain around 0.36V, fed from a 14V-to-19V boost.
- Antminer S19 Pro (BM1398): 114 chips across 38 domains (3 chips per domain), each domain near 0.32V, with a 12.6V-to-20V boost.
- Antminer S21 / T21 (BM1368): 108 chips across 12 domains (9 chips per domain) at roughly 1.2V, boosted toward ~25V, with no PIC chip on the board.
- Antminer S21 XP (BM1370): 91 chips across 13 domains (7 chips per domain) near 1.04V, boosted to ~21V, using 12 level shifters.
This series-domain design is why a hashboard is so unforgiving. Because the chips in a domain share current, one failed chip or one bad regulator can drop the whole domain offline, and a broken link anywhere in the daisy chain can silence every chip downstream of it. That is the difference between a board reading "0 ASIC" and one running at a fraction of its rated hashrate.
Why hashboards fail (and how they get fixed)
Hashboard faults are the single most common cause of lost hashrate on an ASIC miner. The usual suspects are dead or short-circuited chips, burned voltage regulators, cracked solder joints (often from thermal cycling or rough shipping), heatsinks that have fallen off and left chips to overheat, and water or corrosion damage on hydro and immersion units. Because the chips are daisy-chained, a single open circuit in the middle of the string can take out an entire board's reported chip count even when most of the silicon is still good.
Diagnosing this is part detective work, part electronics. Technicians read the chip enumeration the control board reports, then probe the board with a multimeter in diode mode against known-good reference values, hunting for the exact domain or chip where voltage drops out. Repair from there is fine-pitch surgery: hot-air rework to lift and replace a dead chip, reflowing cold solder joints, swapping a failed regulator, or replacing a damaged temperature sensor. It is precise, microscope-and-tweezers work, which is exactly why hashboard repair is treated as its own specialist discipline rather than a casual home fix.
Understanding the hashboard matters because it directly drives uptime and profitability. A miner advertised at a given terahash figure only hits that number when all three boards are fully populated and every domain is holding voltage. When one board drops, you lose roughly a third of your hashrate. Whether you are running a single Bitaxe on the bench or a rack of S21s, knowing how the board is structured tells you what a fault report actually means and whether a machine is worth repairing or retiring.
D-Central has been repairing Antminer hashboards since the early ASIC era, and that bench experience feeds straight back into the firmware and tooling we build. If you have a board reading low or showing dead chips, our team can diagnose and repair it. To learn the surrounding hardware, read up on the control board that coordinates the boards, the ASIC technology behind the chips, and the PCB fundamentals every hashboard is built on. Ready to keep your fleet hashing? Start with our ASIC troubleshooting guides.