Every few months someone in a mining group posts the same photo: a brand new hydro miner sitting on a garage floor next to a bucket, with the caption “how hard can it be.” Two weeks later there is a follow-up post about a leak.
Hydro cooling is genuinely better technology. It is also a different sport. Here is the honest comparison, written for someone deciding where to put real money.
What actually changes when you go hydro
An air-cooled ASIC pushes ambient air across a heatsink with two loud fans. A hydro unit circulates coolant through a cold plate and dumps the heat into a heat exchanger somewhere else. That single change cascades:
- Noise. An air-cooled S21 sits around 75 dB. That is a shop vac running permanently. A hydro unit with the pump and dry cooler outside is closer to 50 dB in the room. This is the single biggest reason people switch.
- Density. You can rack hydro units far tighter because you are not fighting for airflow paths. If you are paying for square footage, this matters more than anything else on this list.
- Overclock headroom. Water carries heat away far better than air, so hydro SKUs ship with higher stock hashrate and tolerate tuning that would thermally throttle an air unit.
- Chip longevity. Stable, lower junction temperatures mean less thermal cycling. Hash boards last longer.
What it costs you
- Plumbing is now your problem. You need a pump, a dry cooler or plate heat exchanger, hoses rated for the temperature, and a coolant loop that does not corrode. None of this ships in the miner box.
- A leak is a total loss. Air-cooled failure mode is “fan dies, unit throttles.” Hydro failure mode can be “coolant on a live PSU.”
- Higher entry cost. The miner is more expensive and the supporting loop is a second purchase of real size.
- Harder resale. The buyer pool for a used hydro unit is much smaller than for a used air miner.
The actual decision rule
Buy air-cooled if: you are running fewer than roughly ten units, you have somewhere to dump noise and heat, or this is your first deployment. The lower complexity is worth more than the efficiency gain at that scale. Most people reading this are in this bucket, and there is nothing wrong with that.
Buy hydro if: you are past ten units, you are paying commercial rent per square foot, you already have someone who can plumb a loop, or noise is a hard constraint (residential, mixed-use, a landlord). The efficiency and density gains compound at scale in a way they simply do not for a three-unit setup.
The middle path most people miss: immersion. Single-phase immersion gets you most of the noise and density benefit with far less catastrophic failure risk than a pressurized water loop, because the dielectric fluid is not conductive. It is worth pricing before you commit to hydro.
Do not skip the power audit
Whichever you pick, run the numbers before you buy. A single modern unit pulling 3,500W continuous on a 240V circuit is not something you improvise. Work out your real cost per kWh, your unit’s J/TH, and what that means at current difficulty. If you want to sanity-check a specific machine, the bitcoin mining profitability calculator will do the arithmetic for you before you spend anything.
Where to look at actual units
If you want to compare real stock rather than spec sheets from a manufacturer PDF, MiningGear Co lists both sides of this: the hydro-cooled ASIC miners sit in their own category, and the conventional air units are under ASIC miners. Comparing the two categories side by side with real prices attached is a faster way to calibrate than reading another comparison table.
The short version
Hydro is better engineering. Air is better logistics. At small scale, logistics wins, and it is not close. At scale, the equation flips hard. Be honest about which one you actually are today, not which one you plan to be in a year.