Equipment & Hardware You'll Actually Need

Almost every crypto mining business plan that fails to raise financing fails for the same reason: the founder priced the ASIC hardware and stopped there, ignoring the electrical, cooling, and racking costs that come with it. A realistic plan prices the whole stack, not just the miner.

Buy outright, or host with someone else's machines?

Before pricing hardware, decide which side of the fence your plan sits on, because it changes every number that follows. Owning hardware outright gives you the full mining upside and lets you choose your own site, but it concentrates obsolescence risk on you - a current-generation ASIC is typically competitive for 18–30 months before efficiency gains from newer silicon erode its margin. Hosting your own machines at a third-party facility (Compass Mining and Core Scientific both offer this) removes the site-selection and electrical-build burden entirely, in exchange for a monthly hosting fee that eats into margin regardless of whether you're mining profitably that month. A growing number of first-time operators split the difference: buy the hardware, but host it at a facility with existing power and cooling, keeping the capital outlay in equipment rather than real estate while still earning on a per-kW basis against an industrial power rate rather than retail.

Mining hardware (ASICs)

Three manufacturers dominate the ASIC market: Bitmain (the Antminer line), MicroBT (the Whatsminer line), and Canaan (the Avalon line, the first mining-hardware maker to go public, on NASDAQ as CAN). As of September 2026, an air-cooled Bitmain Antminer S21 XP (270 TH/s) runs around $4,250, the hydro-cooled S21j XP (495 TH/s) around $9,500, and the newer S23 Hydro line from roughly $13,500 to $19,000 depending on model. At the entry end, a single Avalon Nano or Bitaxe solo miner starts under $200 but produces a negligible, largely symbolic hash rate - fine as a hobby proof-of-concept, not a business.

• ASIC units (20–40 for a small commercial facility): $85,000–$170,000 at current Antminer S21 XP / Whatsminer M60-class pricing
• Power Distribution Units, cabling & breakers: $200–$500 per unit, $8,000–$20,000 for a 20–40 unit fleet
• Cooling (forced-air containment, or an immersion retrofit for high-density racks): $6,000–$45,000
• Racking, fire suppression & noise mitigation panels: $4,000–$18,000
• Fleet management software (HiveOS, Braiins OS+, or Foreman): typically $2–$5 per rig per month, negligible relative to hardware but essential for remote monitoring and firmware control
• Pool connection and failover: a primary pool (Foundry USA, AntPool, F2Pool, or ViaBTC - together controlling close to 70% of network hashrate) plus a secondary pool configured for automatic failover

If the capital outlay for owning hardware outright doesn't make sense yet, hosting providers such as Compass Mining and Core Scientific - one of the largest hosting operators in North America, running more than 600,000 Bitmain units across its data centers since 2017 - will rack and power third-party machines for a per-kW fee. That shifts your plan from a capital-heavy build to an operating-expense model, which is often the more fundable story for a first-time founder.

Cooling: the line item people underbudget

A rack of 20–40 ASICs running at full load is loud and hot in a way that's hard to appreciate until you've stood next to one - individual units commonly run in the 75–90 decibel range at close range, comparable to a vacuum cleaner or light power tool running continuously, 24 hours a day. Most small operators start with simple forced-air containment (ducted intake, hot-aisle exhaust) at the lower end of the cooling budget, then migrate to immersion cooling - submerging boards directly in a dielectric fluid - once density or ambient temperature makes air cooling impractical. Immersion retrofits cost more upfront but cut fan noise dramatically and let you run hardware at a lower junction temperature, which extends the usable life of the board. If your site has residential neighbours within a few hundred metres, price immersion or a sound-attenuating enclosure into the plan from day one rather than reacting to a complaint later.

For day-to-day fleet operation, HiveOS and Braiins OS+ are the two most widely used firmware/fleet-management platforms for monitoring hash rate, temperature, and pool connection across dozens of rigs from a single dashboard, with automatic restart and firmware-level power tuning. Neither is a large line item, but a plan that shows you've chosen one - and budgeted the few dollars per rig per month it costs - reads as more operationally credible than one that treats the fleet as a black box.

What It Costs to Get Mining Online

A small commercial mining operation (10–50 ASIC units) typically requires $50,000 to $250,000 in the US, or roughly £40,000 to £195,000 in the UK, to get from signed lease to first block reward. Hardware is usually the single largest line item, but it's rarely the one that sinks a plan - the electrical and cooling build-out is.

Cost breakdown (20–40 unit facility)

• ASIC hardware (Antminer S21 XP / Whatsminer M60 class): $85,000–$170,000 (£67,000–£134,000)
• Electrical upgrade, PDUs, cabling & breakers: $8,000–$35,000 (£6,000–£28,000)
• Cooling - forced-air containment or immersion retrofit: $6,000–$45,000 (£5,000–£35,000)
• Facility lease deposit & light-industrial fit-out: $10,000–$40,000 (£8,000–£32,000)
• Noise mitigation, fire suppression & racking: $4,000–$18,000 (£3,000–£14,000)
• Working capital - 3 months of electricity & pool fees: $7,000–$30,000 (£6,000–£24,000)

Funding routes

This is the single biggest way a mining business plan differs from almost every other industry template on this site: standard small-business lending generally will not touch it. The SBA explicitly excludes speculative businesses from 7(a) eligibility, and most lenders classify crypto mining under that heading because revenue is a function of a volatile asset price and shifting network difficulty, not a predictable unit sale. The UK's Start Up Loans scheme faces the same practical issue - mining doesn't fit the standard affordability assessment most delivery partners run.

In practice, operators fund hardware through vendor financing direct from Bitmain or MicroBT, equipment-backed loans secured against the ASICs themselves, hosting contracts that convert capex into opex, or private investor capital. A plan written for a crypto mining business needs to lead with this reality rather than assume a bank loan is on the table - our bespoke business plan service builds the financing narrative around whichever of those routes actually fits your situation.

A typical vendor-finance structure looks like this: 20–30% of the hardware cost as a deposit, the remainder amortised over 12–24 months at a fixed rate, with the ASICs themselves standing as collateral so the lender's downside is the resale value of the machines, not your business's cash flow. Equipment-backed lenders will usually want to see at least three months of operating history or a signed hosting contract before they'll quote terms, which is why many first-time operators start with a small self-funded pilot - 4 to 10 units - purely to generate the track record a larger financing round will later ask for. Build that staging into your plan's timeline rather than assuming the full fleet goes live on day one.

Where You Mine Matters: A Regional Map

Unlike a daycare or a restaurant, a mining business can genuinely choose its jurisdiction, because the product - hash power - doesn't need to be near a customer. That makes the regional decision one of the most consequential choices in the whole plan.

A quick glossary before the regional detail, since these three terms drive almost every number in this section: hash rate is the number of calculations your hardware (or the whole network) performs per second, measured in TH/s per machine and EH/s network-wide; difficulty is a network-wide setting that adjusts roughly every two weeks so that, regardless of how much total hash rate joins the network, a block is still found on average every 10 minutes - more hash rate joining the network pushes difficulty up and reduces everyone's individual share of the reward; and the halving is a programmed event, roughly every four years, that cuts the block reward in half - the most recent cut it to 3.125 BTC per block, with the next one expected around 2028 to bring it to 1.5625 BTC. Every revenue projection in a mining plan is ultimately a bet on how those three numbers move relative to the price of Bitcoin.

United States

• Texas: the most mining-friendly large grid in the US. ERCOT runs a voluntary curtailment program for "Large Flexible Loads," paying miners to power down during demand spikes - in August 2023 alone, ERCOT paid one Bitcoin miner $24.2 million in power curtailment credits plus a further $7.4 million through its separate demand response programme. That's real, bankable revenue on top of mining income, not a theoretical incentive.
• Kentucky: offers sales and use tax exemptions on mining equipment and electricity for qualifying facilities, plus some of the cheapest industrial power in the country.
• New York: enacted a two-year moratorium in 2022 on new proof-of-work mining operations powered by non-renewable energy - a constraint that is still shaping site selection for new entrants.
• California: permits mining but layers on strict zoning rules, high electricity costs, and - from 1 July 2026 - new licensing under the state's Digital Financial Assets Law that crypto-adjacent businesses should confirm applies (or doesn't) before signing a lease there.

Outside the US

Kazakhstan became one of the world's largest mining hubs after China's 2021 ban pushed miners out, but it since introduced a mining-specific levy: operators must register with the state digital-asset regulator and pay a progressive electricity tax tied to the spot price of power consumed for mining, on top of standard industrial tariffs - a cost most other jurisdictions simply don't impose. If your plan assumes Kazakhstan-style cheap power, it needs to also assume that tax.

Two things matter more than headline electricity price when you're comparing sites: the utility's interconnection queue and the power mix. A site with cheap power on paper is worthless if the utility tells you a new 2–5MW connection won't be energised for 12–18 months - which is increasingly common in grid-constrained parts of Texas, Georgia, and Pennsylvania as data-centre demand competes for the same capacity. Build the interconnection timeline into your launch plan explicitly, with a fallback site if the primary connection slips. On power mix, Alberta in Canada has become a secondary North American hub worth comparing against Texas: deregulated power pricing and a grid operator (AESO) that, like ERCOT, has started experimenting with demand-response products for large flexible loads.

Further afield, Sweden and Norway attract miners for cheap, largely hydro-based power and a cold climate that cuts cooling costs significantly compared to Texas or Kazakhstan - the tradeoff is a higher cost of labour and construction, and in Sweden specifically, periodic political pressure to tax or restrict mining's share of renewable grid capacity. Any plan sited in the Nordics should treat the heat-reuse angle (selling waste heat from the racks into a district heating network) as a genuine, fundable revenue line rather than a sustainability footnote - several Nordic data-centre operators already have commercial heat-offtake agreements in place, and the same model applies directly to a mining facility's exhaust heat.

For a UK-based founder weighing US siting against staying closer to home, our free business plan template includes a jurisdiction-comparison worksheet you can adapt to whichever two or three locations you're actually evaluating.

Licenses, Registrations & Compliance

The most common mistake in this section of a crypto mining business plan is treating "mining" and "running a crypto financial service" as the same regulatory category. They aren't, and lenders and investors notice when a plan conflates them.

United States

• Proprietary self-mining (mining for your own wallet) is generally exempt from FinCEN money-services-business registration under FinCEN's 2019 guidance
• Hosting other people's rigs, selling hash power, or converting mined coin to fiat on a customer's behalf typically does require FinCEN MSB registration and, in many states, a money-transmitter license
• Check your state against New York's proof-of-work moratorium and California's incoming DFAL licensing before committing to a site
• Utility interconnection agreement for any load large enough to need a dedicated supply - in Texas this means formal registration as a Large Flexible Load with ERCOT
• Standard business licensing, zoning approval for industrial/data-center use, and fire code compliance for high-density electrical racks

United Kingdom

• Pure mining is not itself an FCA-regulated activity
• Since 2021, any cryptoasset business with AML-relevant activity - hosting, exchange, or custody services layered on top of mining - must register under the Money Laundering Regulations 2017 with the FCA
• The FCA opened its full cryptoasset authorisation window on 30 September 2026, running to 28 February 2027, ahead of the full regime taking effect on 25 October 2027. Firms already MLR-registered do not automatically carry over - reapplication is required
• A commercial electrical connection agreement with your local Distribution Network Operator, since mining loads routinely exceed standard light-industrial allowances
• Local authority planning consent for industrial/data-center use class, plus noise assessment where racks will run near residential boundaries

In practice, UK local authorities treat a mining site similarly to any other noisy light-industrial tenant: expect to be asked for a noise impact assessment if you're within roughly 200 metres of residential property, and expect continuous rig noise in the 75–90dB range at source to fail most residential-boundary noise limits without an acoustic enclosure. For company formation itself, registering with Companies House and setting up PAYE/VAT registration with HMRC is standard regardless of the cryptoasset-specific steps above - don't let the crypto-specific compliance checklist distract from the ordinary company-formation one sitting underneath it.

Firms that do trip the AML-registration threshold - typically because hosting, custody, or hash-power resale is part of the business model - will also need to name a Nominated Officer responsible for anti-money-laundering compliance and suspicious-activity reporting, carry out customer due diligence on anyone paying to host or buy hash power through you, and keep records for the FCA's minimum retention period. Build the cost of that compliance function, even if it's a part-time role in year one, into the operating budget rather than discovering it during the authorisation application.

Tax treatment (both markets)

The IRS treats mined cryptocurrency as ordinary income, valued at fair market value on the date received - not the date you eventually sell it. If mining is run as a genuine business rather than a hobby, it's reported on Schedule C, which makes it eligible for deductions on equipment, electricity, and hosting fees, but also adds a 15.3% self-employment tax on net profit. HMRC takes a similarly literal view in the UK: mined coin is taxable as income (or as a company's trading profit) at the point it's received, with capital gains rules applying separately when it's later disposed of. Get an accountant who has actually filed a mining return before you finalize the financial model - this is not a section to estimate.

How Mining Businesses Actually Make Money

Unlike most businesses on this site, a mining operation doesn't set its own price. Revenue is simply your share of the network's block rewards, multiplied by whatever Bitcoin happens to be worth that day. That reframes the entire plan: the variables you control are hardware efficiency, electricity cost, and uptime - not price.

Three commercial models

• Self-mining: you own the hardware, you own or lease the site, and you keep 100% of the mined coin minus pool fees (typically 0–2%) and electricity
• Hosting / colocation: you rent out rack space and power to third-party rig owners for a per-kW fee, usually priced as your electricity cost plus a markup - Compass Mining and Core Scientific both run this model at scale
• Hash-power resale / cloud mining: you sell slices of your hash rate to retail customers who never touch physical hardware - the highest-margin model on paper, but the one most likely to trigger FinCEN MSB or FCA registration, because you're now handling other people's money, not just your own

Worked example

Take a 40-unit Antminer S21 XP fleet: 270 TH/s per unit, 10.8 PH/s combined, drawing roughly 3.5 kW per unit - about 140 kW continuous for the whole fleet. Hosted at $0.065/kWh, that's approximately $6,550 a month in electricity. Using an approximate September 2026 network hashrate in the 900 EH/s range and the current 3.125 BTC block reward, that fleet's share of the network works out to roughly 0.16–0.17 BTC per month. At a BTC price near $86,300, that's somewhere around $13,800–$14,800 in monthly gross revenue, leaving a pre-overhead margin in the same 45–55% band the sector was reporting in September 2026 for sub-15 J/TH hardware. This is a labelled, composite illustration, not a quote - both the Bitcoin price and network difficulty move constantly, and your plan's financial model needs a downside case built on at least a 30–40% price drop, not just the current spot price.

Additional revenue some operators layer on top: grid curtailment payments (see the Texas ERCOT example above), waste-heat resale to adjacent greenhouses or district heating in cold climates, and transaction-fee upside during periods of high network congestion, which can meaningfully exceed the base block subsidy during busy weeks. During sustained congestion, aggregate transaction fees have historically spiked to represent a noticeably larger share of total miner revenue than the roughly low-single-digit percentage they contribute in a typical week - a genuine, if unpredictable, upside case worth a sentence in your plan's revenue section, not a line item you should ever treat as reliable base-case income.

If you're building the hosting model rather than self-mining, structure the fee the way the established hosting operators do: a flat rate per kW of IT load (commonly in the $0.08–$0.12/kWh all-in range once your margin is layered over the raw electricity cost), billed monthly, with a minimum-term commitment of 6–12 months so your own revenue isn't exposed to a customer pulling their machines the week after you've committed to a lease. A business plan that only models the self-mining case and ignores hosting as a second revenue stream is leaving a genuinely investor-relevant diversification argument on the table.

Five Mistakes That Sink a First-Time Mining Plan

• Signing a lease before confirming power capacity. Utility interconnection queues for a meaningful commercial load can run 6–18 months. Founders who sign a facility lease first and apply for the power upgrade second often end up paying rent on a site they can't yet energise.
• Budgeting electricity at a residential rate. The gap between a residential tariff (~$0.17/kWh in the US) and a negotiated industrial or demand-response rate (~$0.05–$0.07/kWh) is the single biggest swing factor in the entire financial model - get a commercial quote before you finalise projections, not after.
• Ignoring noise and zoning until neighbours complain. A site that passes on paper can still get shut down by a local noise complaint or a zoning challenge months into operation if acoustic mitigation wasn't priced in from the start.
• Modelling revenue only at today's Bitcoin price and today's network difficulty. Both move constantly, usually against the miner - difficulty has historically trended upward as more hash rate joins the network. A credible plan includes a downside case built on at least a 30–40% price drop and a meaningful difficulty increase, not just the optimistic spot-price scenario.
• Treating a hosting or hash-power-resale offering as "just mining." The moment you're handling other people's hardware or hash rate for a fee, you've likely crossed into FinCEN MSB or FCA registration territory. Plans that skip this distinction tend to get flagged hardest by investors who've seen a prior mining deal get tripped up by it.
• Leaving Bitcoin price risk completely unhedged. Some operators sell mined coin immediately to lock in fiat revenue and fund operating costs predictably; others hold some portion as a treasury asset, the way MARA Holdings and Riot Platforms do at scale. Either is a defensible strategy - but a plan that never states which one it's following, and why, leaves a lender or investor to assume you haven't thought about it.

The Market in Numbers

The global cryptocurrency mining market was valued at approximately $26.89 billion in 2025, rising to an estimated $31.76 billion in 2026, according to Precedence Research. A narrower scope focused specifically on mining services alone - excluding hardware manufacturing - puts 2025 at closer to $4.61 billion per Grand View Research. Both are directionally right; the gap is scope, not disagreement - one counts the hardware supply chain and hosting revenue together, the other counts mining-service revenue alone. Build your plan's market sizing on whichever definition matches what you're actually selling.

Who's winning

The publicly traded mining sector is a useful proxy for where capital and operational discipline are flowing. By market capitalisation in 2026, the largest listed miners are IREN (~$13.8B), Hut 8 (~$10.6B), TeraWulf (~$8.1B), Riot Platforms (~$7.3B), Cipher Mining (~$6.8B), Core Scientific (~$6.2B), MARA Holdings (~$3.7B), and CleanSpark (~$3.0B), according to Capital.com's 2026 ranking. By Bitcoin held rather than market cap, MARA Holdings leads with roughly 38,689 BTC, ahead of Riot Platforms (~18,005 BTC) and CleanSpark (~13,099 BTC). None of these companies compete with a small operator directly - but their public filings are a genuinely useful, free source of real-world cost and efficiency benchmarks when you're building your own financial model.

Two structural trends worth knowing before you write the plan

First, hash rate has kept migrating toward cheap, often stranded or curtailable power - the same pattern that pulled the industry out of China after its 2021 mining ban and into Texas, Kazakhstan, and parts of the Nordics. A new site plan that doesn't address why its power is cheap (stranded gas, demand response, hydro overcapacity, a negotiated industrial tariff) reads as less credible to anyone who's underwritten a mining deal before. Second, several of the largest public miners - Core Scientific and IREN among them - have been converting a meaningful share of their data-centre capacity toward AI and high-performance-computing hosting rather than pure Bitcoin mining, chasing more stable, non-crypto-denominated revenue. That's a scale-driven pivot that doesn't change the economics for a small operator today, but it's a useful signal of where the industry's largest, best-capitalised players see the next few years of risk sitting.

Try It: Rough-Cut Profitability Calculator

This is a simplified estimate, not a financial model - but it's the same basic math a lender or investor will run on the back of an envelope before they read your plan. Adjust the defaults (pre-filled to roughly match a 40-unit Antminer S21 XP fleet) to your own numbers.

Formula: daily BTC = (your TH/s ÷ network EH/s×1,000,000) × 144 blocks/day × 3.125 BTC block reward. This ignores transaction-fee upside, difficulty changes, and downtime - all three move the real number. Use it as a sanity check, not a forecast.

Inside a Real Mining Business Plan

Here's an extract from a mining business plan written by our team - so you can see exactly what you'll get. Note how the financial section leads with the downside case rather than burying it in an appendix; that's the single biggest structural difference between a plan that gets funded and one that gets politely declined.

Everything That Ships Inside the Template

We built this template specifically for mining operators rather than adapting a generic small-business template, which is why it leads with hardware and power economics instead of a conventional retail-style marketing plan. Every Avvale business plan template includes these sections, pre-structured for your industry:

• Executive Summary - Your mining operation at a glance, written to hook a lender or investor in 60 seconds
• Company Overview - Legal structure, hardware ownership model, site, and founding story
• Industry Analysis - Market size, network difficulty trends, and the regulatory landscape across your target jurisdictions
• Hardware & Infrastructure Plan - ASIC selection, power and cooling design, hosting vs. self-owned tradeoffs
• Competitor Analysis - Public-miner benchmarking and your differentiation strategy against hosting alternatives
• Operations Plan - Uptime targets, pool failover, maintenance schedule, and key milestones
• Financing Strategy - Vendor financing, equipment-backed lending, and investor-capital structuring, written around mining's SBA-ineligibility rather than ignoring it
• Management Team - Founder bios, technical advisors, and key hires planned

The optional Financial Forecast add-on (included in our $300/£250 and $1,000/£800 packages) provides a 5-year Excel model with a base case, a downside case stress-tested against a 30–40% BTC price drop, hosting-fee sensitivity, and a difficulty-growth curve - not just a single static projection at today's price.

Questions Founders Ask

Related Business Plan Templates

If a boutique rig count isn't the goal, our bitcoin mining farm business plan template covers the larger-scale, multi-megawatt version of this plan. If you're exploring adjacent digital-asset infrastructure rather than mining itself, our crypto ATM business plan template walks through a very different regulatory and revenue picture worth comparing against this one before you commit capital to either.