How a Bitcoin Miner Became a Smart Water Heater

I work in the crypto self-custody space. In late January 2026, Winter Storm Fern knocked roughly 40% of US hash rate offline. Around 455 EH/s went dark. Foundry alone lost 60% of contributed hash, and CoinDesk reported a 12% price drawdown, the worst since the China ban. The ASICs did not fail. The homes they were installed in needed every available watt for warmth. That sentence is the future of Bitcoin mining.

This piece builds on Tyler Stevens' interview on the Bitcoin Rails podcast. Tyler wrote Bitcoin Mining Heat Reuse (Braiins, 2025), sits on the board of the 256 Foundation (501(c)(3), EIN 99-1662333), runs Exergy Heat as CEO, and founded Hashrate Heatpunks. Watch the long version below in his own voice, then come back for the 2026 reader's guide.

TL;DR

• A Bitcoin miner is a 100%-efficient electric heater that pays you in sats.
• Industrial mining hit a ceiling: closed firmware, three-phase power, no heat offtaker.
• The next wave runs in homes, water heaters, and small commercial buildings.
• The math works best in cold-climate rural homes burning propane or heating oil. Natural gas pipelines stay on natural gas.
• The bottleneck is open-source hardware, not physics. Bitmain holds roughly 40% of the ASIC market (35 to 45% range, SkyQuest/FutureMarket 2024-2025). Bitmain plus MicroBT control roughly 70%.
• The 256 Foundation is shipping an open stack: Ember One hashboard, Mujina firmware, Libre Board control board, Hydra Pool mining pool.
• In 30 years, most of Bitcoin's hash rate may not run on Bitcoin mines at all. It will run on appliances.

1. The Old Story Was Wrong: Bitcoin Miners Are 100% Heaters

Drop a Bitcoin miner into a closed room and measure the temperature. It rises. The energy has nowhere else to go. That is not "waste."

A Bitcoin miner is an electric resistive heater. Power flows through silicon. The silicon resists. The resistance produces heat. Same physics as your toaster, your kettle, and the baseboard in your guest bedroom. The one difference: the silicon also tries to guess a number that, if correct, mints a block reward.

The "wasteful mining" narrative assumed the heat had nowhere to go. That assumption broke quietly between 2020 and now.

Reframe it as heater that pays you in sats instead of miner that emits waste heat. That is the pitch of the home mining movement. Once you see it, the rest of this article is engineering and money.

2. The Industrial Ceiling: Why Mega Miners Are Cornered

Walk into a Texas mining warehouse. Twenty thousand machines. 480-volt three-phase service. Fans the size of small cars. Ambient temperature around 95 degrees F. The heat is enormous, and it vents to the desert sky.

That is the mega-miner model: get the cheapest electricity, run the latest ASIC, treat heat as a problem to engineer around. It worked while Bitcoin was small and electricity was cheap. It now hits three structural limits.

Three-phase power locks them out of the home. MicroBT's new M70-series lineup targets industrial three-phase service for its higher-tier hydro and immersion models. That is not a household outlet. It is not even a dryer outlet. It is dedicated industrial service. Mega-miners run away from residential faster than residential can run toward them.

Closed firmware is a single point of failure. Bitmain holds roughly 40% of the ASIC market (35 to 45% range). Their firmware is closed-source. We have caught them before. Antbleed, a remote kill-switch baked into older firmware, was disclosed April 26, 2017 via antbleed.com (CoinDesk). Greg Maxwell publicly disclosed ASIC Boost, a chip-level optimization Bitmain allegedly used internally, on April 5, 2017. Every mega-miner trusts a single Chinese company not to skim from the top, brick the fleet, or sell the hash rate out to a state actor.

Hash price keeps grinding toward zero. As more efficient machines flood the network, the dollar value of one terahash falls. Mega-miners survive by being the lowest-cost producer. The whole industry runs a treadmill toward zero margin. The bottom of that hill is hash price = 0. At that point, the only mining that survives is mining where the heat is the product.

That is the moment the industry pivots. The megas did not choose it. They got cornered.

3. The Heat Reuse Stack: Where Mining Goes Next

Home mining is not one product. It is a stack of four, each with its own physics, economics, and customer.

3a. Space heaters: the gateway drug

This is where the movement started. Hobbyists at meetups, an old Antminer S9, a Loki kit converting the power supply to 110V household current, larger Noctua fans replacing the screaming stock fans, a 3D-printed shroud. You end up with a 1.5 kW Bitcoin space heater in a corner of your living room.

Pros: cheapest entry point, fully reversible, you tinker with it yourself. Cons: low duty cycle (you only run space heaters when it is cold), still loud unless you put real money into the retrofit.

3b. Water heaters: the sleeper hit

Almost no one talks about water heaters. They should. A hot-water tank is a closed loop. You heat 50 gallons once and the heat sits there until someone takes a shower. The element runs at low duty cycle, but the heat demand stays constant across the year, summer included.

A Bitcoin water heater wants the opposite of a Bitcoin space heater. A tiny miner running long hours, not a big miner running occasionally. That flips the product design. Recovery time is the engineering headache (a tiny miner takes forever to reheat a drained tank). The addressable market is every house in the country, every season.

3c. District heating: the EU advantage

Marathon Digital launched a 2 MW pilot in Satakunta, Finland in May 2024, heating roughly 80,000 residents. A second project targets a city of 67,000 (datacenterdynamics.com). This is not a wound-down experiment. It is active and expanding. District heating is rare in the US and standard infrastructure in Northern Europe and Iceland. That asymmetry alone could put EU heat-reuse mining years ahead.

3d. Industrial low-grade heat (below 180 degrees F)

Bitcoin miners cap out around 180 degrees F before the silicon dies. That puts them in the "ultra-low-grade heat" bucket. Useless for steel, glass, or chemicals. Perfect for:

• Greenhouse heating in northern climates (Gridless Compute Africa is pioneering productive-use models across the continent)
• Brewery and distillery operations
• Truck-wash bays in mountain states (a friend of Tyler's runs one in Idaho)

These applications run high duty cycle, weather-independent, and the business already pays for the heat. Replacing the propane bill with a Bitcoin heater is a straight P&L upgrade.

4. The Money: When Does a Bitcoin Heater Pay Off?

Nobody on Bitcoin Twitter wants to run this comparison honestly. Convert every heating fuel to a $/kWh equivalent, then run the subsidy math.

At hashprice $39/PH/day (Luxor hashrate index, 2026-05-11), an Antminer S19 at 95 TH/s earns roughly $3.71/day. Grid electricity at $0.18/kWh (EIA Feb 2026 actual: 17.65 cents; EIA STEO May 2026 forecast: 18.2 cents) and 3.25 kW x 24h = 78 kWh/day puts the bill at roughly $14/day. You get a 26% offset on US grid power. On rural propane or heating oil the offset climbs to 30 to 50%.

Sources: EIA STEO May 2026, EIA weekly retail prices Feb 2026, Luxor hashrate index 2026-05-11, Bitmain S19 spec sheet (95 TH/s, 3,250W). Natural gas $1.00 to 1.50/therm divided by 29.3 kWh/therm. Heating oil note: Arctic Village AK reached $15/gal ($0.38/kWh equivalent) in winter 2025-2026.

Your bullseye customer is a rural household in a long, cold heating season burning propane or oil. Cities with natural-gas pipelines stay on natural gas, for now. Florida and Phoenix never enter the conversation.

Three business models for getting Bitcoin heaters into homes

1. Own and hold. Customer pays full price for the heater, keeps the heat, keeps the sats. Highest upfront cost, simplest psychology. Works today.
2. Revenue split. Customer pays a discount. The installer, utility, or HVAC company keeps a slice of the mining rewards. Aligns incentives across the supply chain. Needs new contracting infrastructure.
3. Free heater, company keeps the sats. The radical model. The manufacturer subsidizes the hardware to near-zero and bets on years of mining revenue from millions of installed units. Works only when you can guarantee uptime, so it fits water heaters and furnaces (no one unplugs those), not portable space heaters.

None of these models has been tested at scale. The first to crack the "free heater" model probably becomes the largest Bitcoin miner in the world by accident.

5. The Bottleneck: Open-Source Hardware

Everything above is real. None of it scales while Bitmain plus MicroBT control roughly 70% of the ASIC market. For the protocol mechanics behind this (proof-of-work, ASIC economics, block subsidy and fee interaction), see the technical deep dive chapter.

This is not an ideological complaint. It is an engineering one. To build a Bitcoin water heater you need ASIC chips. To buy ASIC chips at scale you need a relationship with Bitmain or MicroBT. They sell to the highest bidder, currently CleanSpark, Marathon, and Riot, not to a startup pivoting the heating industry. The open-source efforts (Bitaxe and friends) desolder chips out of used Antminers shipped from China. The Bitaxe community originally salvaged BM1397 chips from S17 units. That is not a manufacturing strategy.

The fix is the 256 Foundation, a 501(c)(3) nonprofit (EIN 99-1662333) building a fully open-source mining stack across four projects:

• Ember One: open-source hashboard
• Mujina: open-source firmware (so you can verify what your machine does)
• Libre Board: open-source control board
• Hydra Pool: one-click open-source mining pool with multiple payout structures

In January 2025 they ran Telehash #1: peak pointed hashrate of approximately 1.12 EH/s from roughly 350 entities. The foundation found a block when approximately 881 PH/s was pointed (nobsbitcoin.com). As of last month, the reference design, control board, hashboard, firmware, and pool are all open source and mining live. The remaining gap is the chips themselves. Proto (Block, Jack Dorsey's hardware company) has confirmed pursuit of open silicon. Getting competitive chips into production needs tens to hundreds of millions in commitments and TSMC access. A competitive open-source ASIC is a 5 to 10 year project, if it ships at all.

The pragmatic goal is not "open-source chips." It is "more than one chip vendor that will sell to you." That alone breaks the residential mining bottleneck.

6. Should You Mine at Home in 2026? A Decision Framework

Run these six questions before you spend a dollar on hardware:

1. Climate. How many heating-degree-days per year? Below ~3,000, the duty cycle will not carry the math.
2. Primary fuel today. Natural gas? Stop here. You will not beat $0.04/kWh equivalent. Propane, heating oil, or grid electric? Continue.
3. Electricity rate. Below ~$0.13/kWh, the subsidy works cleanly. Above ~$0.20/kWh (parts of the EU and California), the offset shrinks.
4. Noise tolerance. Do you have a basement, a utility closet, or a detached garage? If your only option is the living room, run a Bitaxe Gamma and read the rest of this for context.
5. Time budget. A retrofit takes a weekend. Hydronic integration takes weeks. Plug-and-play heatpunk products are still emerging.
6. Motive. Sats? Bitaxe at home for the education. Heating bill? Run the math against your fuel cost. Sovereignty? Build the open-source stack and point a Bitaxe Gamma at Hydra Pool.

Three tiers, three real builds

• Beginner: Bitaxe Gamma. Around 15W (less than your phone charger), BM1370 chip, single open-source ASIC, pre-built and shippable. Mines about as much heat as a hand-warmer. The point is not the heat. You personally run a miner without trusting Bitmain. Source: bitaxe.org, github.com/skot/bitaxe.
• Intermediate: retrofit Antminer S19 with a Loki kit. Around 3 kW, real space heater, ~50 dB once you swap the fans. Plan for a dedicated 30A circuit and acoustic isolation. Loki kits ship from D-Central (d-central.tech), Altair Technology (altairtech.io), Nakamoto Mining, and Pivotal Pleb.
• Advanced: hydronic / hot-water integration. Heatbit Trio (400W mining + 1100W heat output, ~40 dB), Heatbit Maxi (39 TH/s, 1500W), or DIY immersion-cooled S19s tied into a hot-water loop. Plumber-and-electrician territory. The Bitcoin heater replaces an existing appliance line item.

What NOT to do

Do not sign up with a hosting company. The pitch is "Bitcoin mining is too hard, we will run it for you and mail you the sats." That is the modern equivalent of leaving your coins on Mt. Gox. If you cannot see the machine, you do not own the hash rate. The point of self-custody is the same point here: hold the keys, hold the hash, hold the heat.

7. The Long View: 30 Years Out

Tyler's prediction: in 30 to 40 years, the vast majority of Bitcoin's hash rate is not owned by Bitcoin mining companies. It is owned by appliances. Water heaters, furnaces, district-heating loops, server-room HVAC, brewery boilers, every gadget you already pay to run, with hash rate baked in at the factory.

Mining stops being a separate business. It becomes a utility function, the way a refrigerator's compressor is not a "refrigeration startup" but a feature. At that scale, hash price can fall to zero and the network stays maximally secure. The mining is not financed by the block reward. It is financed by your hot water bill.

Thought experiment: General Electric, Bosch, or Vaillant becomes the largest Bitcoin miner on the planet by accident, after shipping 50 million heaters with hash boards inside. The mining itself does not move GE's P&L. It is a free option on top of a heater they were going to sell anyway.

If that future arrives, the Bitcoin map of the world becomes a hash-rate heat map of the world. Highest hash density in the coldest, longest winters. The next cold front in the US Midwest does not dip hash rate, it spikes it.

Whether that future arrives at all turns on the open-source hardware bottleneck above. The fundamentals are good.

About the author

Mohamed Habbat wrote Bitcoin: Zero to Hero, a 19-chapter book that takes complete beginners to functioning Bitcoin self-sovereignty. He works in the crypto self-custody space and has spent the last five years inside Bitcoin.

8. Frequently Asked Questions

Further Reading

• Bitcoin self-custody guide: hold your own keys before you hold your own hash rate
• Bitcoin halving explained: why the block subsidy schedule shapes mining economics
• Bitcoin tax guide (Switzerland): how mining income is treated under Swiss law
• Technical deep dive chapter: proof-of-work, ASIC economics, and block subsidy mechanics