How do you calculate realistic ASIC profitability long term?

[MiningIntel]

I'm curious how experienced miners estimate long-term ASIC profitability.

Most calculators show profitability using today's BTC price and network difficulty, but obviously those numbers don't stay constant.

When deciding whether to buy an ASIC, which factors do you consider most important: electricity cost, J/TH efficiency, difficulty growth, BTC price, hardware price, or something else?

Do you calculate several scenarios (bear/base/bull), or mostly focus on current profitability?

Interested to hear how people who actually operate miners approach this.

[stompix]

There is no universal response for this and there never will be.
It comes down to your own profile, from money available to the electricity available in both price and quantity.

You have a fuckton of cheap 0.01c/kwh that you can feed an aluminium power plant, but it's a deal you might only have for a year and you're strapped on cash? You go and buy the cheapest ASIC on the marketplace that will yield you the most in one year for the money spent, sure, the 19xp can only get you $5 a day compared to the $30 for the s23u but you can get 30 of them used for the price of an S23U.
You have 0.01c/kwh but only 10KW of power available but you know you will have this like forever? Then you simply squeeze the best you can do in that thing, as 30s19 will demand 100KW and you can only fit in either 2 old-generation or 3 older than that, or a brand new one.
You pay 10cents/kwh? You go out in the city , have a drink and forget everything about this!
Then, there comes the problem of the money you have to spend, because it's one thing if you have $1000 or $100k or you have $500k of investors' money you can afford to lose. 

Extreme scenarios but in short, it's the same as a question like,  I have some land, what should I plant?

[philipma1957]

calculation is meh.

factor 1 power cost
factor 2 power amount
factor 3 power cost and amount staying stable for how much time.


if you mine in a home unless you have solar energy the power is likely not cheap enough to mine for profit.

ie mine solo mining for hobby and bragging rites if you ever hit a block.

Now lets say your cost is 10 cents to 20 cents a kwatt and you need some heaters in the winter

if true you could buy a used unit and run it down clocked for the winter months.


Lets say your power is over 20 cents forget mining except for a tiny solo miner like this

https://altairtech.io/product/hammer-miner-bc01-home-bitcoin-miner/


now if you want to mine at commercial levels and do not have cheap power you need a host.

1) first rule for the host is he honest
2) second rule for the host is he honest
3) cost of power
4) cost of power.

if you want to mine at an honest spot my guy has been good for 2 years.

Nate at corex

https://www.corexhosting.io/

this gear is $3550 270th

https://www.corexhosting.io/product-page/bitmain-s21xp-270t

this is hydro cooled gear 6900 550th
https://www.corexhosting.io/product-page/whatsminer-m73s-hydro


now if you want to go professional and buy from above I will tell you it is marginal

the top unit burns 3600 watts an hour or 3.6x732=2,635.2 kwatts a months and power cost is about

243 usd the unit earns about 328.32 a month so 85 a month paid back towards the $3550 price of the gear

that is marginal a strong chance you will not turn a profit.

the second unit is about 2x on power and earnings which makes it very marginal

there is a strong chance you will not benefit from them.

mining is hard real hard.

it is not fast money.

[FP91G]

I'm curious how experienced miners estimate long-term ASIC profitability.

Most calculators show profitability using today's BTC price and network difficulty, but obviously those numbers don't stay constant.

When deciding whether to buy an ASIC, which factors do you consider most important: electricity cost, J/TH efficiency, difficulty growth, BTC price, hardware price, or something else?

Do you calculate several scenarios (bear/base/bull), or mostly focus on current profitability?

Interested to hear how people who actually operate miners approach this.

I've been using a very simple method for over 10 years.
Open any calculator, enter your ASIC and electricity costs, and see the payback period - say, 14 months. My recommendation is to multiply this period by 2. If you have a halving during this period, then increase the payback period proportionally to the number of months after the halving. If after the halving you have 8 months left to mine using the old calculation, then you should calculate 16. In a good scenario, you can break even within a year, but your business plan should be considered for a worse-than-average scenario.

[MiningIntel]

That's a really interesting approach. I especially like the point about adjusting the payback period around the halving rather than relying on today's numbers.

Do you also factor in difficulty growth separately from the halving, or do you mostly use the 2x payback rule as your safety margin?

[FP91G]

That's a really interesting approach. I especially like the point about adjusting the payback period around the halving rather than relying on today's numbers.

Do you also factor in difficulty growth separately from the halving, or do you mostly use the 2x payback rule as your safety margin?

If you're choosing ASICs, the ideal option is one with a payback period of 12-15 months or less. If there's no halving in the next two years, you can skip the complicated calculations and assume you'll recoup your equipment or even make a profit within two years. Since the halving is 20 months away, and you bought an ASIC with a payback period of 12 months, you should adjust your calculations by 4 months. This will assume a payback period of 28 months instead of 24.

[FP91G]

Separately, and it stacks with the halving rather than replacing FP91G's rule. His 2x heuristic is pricing in the halving's one-time reward cut. Difficulty growth is a different mechanism, continuous erosion of your share of the network, happening every two weeks whether or not a halving is anywhere nearby.

Ran it on philipma1957's own numbers from earlier in this thread, $3550 unit, $85/month at today's price and difficulty. Zero difficulty growth, pure static math, that's 42 months. Add 15% difficulty growth a year, roughly a normal year historically, and it's 57. Add 30%, closer to a fast year, and it's 107, nine years, on a unit he already called marginal at today's numbers with no growth assumed at all.

So if you're stacking both effects: take FP91G's halving-adjusted payback, then run difficulty growth on top of that result, not instead of it. A unit that looks like a 28 month payback after the halving adjustment could be pushing 40+ once difficulty is doing its normal thing on top.

A 40-month payback period is a likely loss.
I'm increasing the ASIC's payback by 2, and I'm already factoring in the increased difficulty and the negative scenario, but let's be honest, it's very difficult to make calculations for the next 24-30 months in the Cryptocurrency market. Miners have been lucky in the past, and all calculations are based on past performance.

MiningIntel, tell me your ASIC, its price, and the electricity cost. I'll do my calculations, and we can laugh together in a couple of years.

[philipma1957]

Separately, and it stacks with the halving rather than replacing FP91G's rule. His 2x heuristic is pricing in the halving's one-time reward cut. Difficulty growth is a different mechanism, continuous erosion of your share of the network, happening every two weeks whether or not a halving is anywhere nearby.

Ran it on philipma1957's own numbers from earlier in this thread, $3550 unit, $85/month at today's price and difficulty. Zero difficulty growth, pure static math, that's 42 months. Add 15% difficulty growth a year, roughly a normal year historically, and it's 57. Add 30%, closer to a fast year, and it's 107, nine years, on a unit he already called marginal at today's numbers with no growth assumed at all.

So if you're stacking both effects: take FP91G's halving-adjusted payback, then run difficulty growth on top of that result, not instead of it. A unit that looks like a 28 month payback after the halving adjustment could be pushing 40+ once difficulty is doing its normal thing on top.

but you need to remember
difficult has long downward streches

June 8 2025 we were 127t
Nov 6 2025 we were  155t
Sept 1 2026 we are   126t

so where did difficulty rise in those months

as compared to other time periods.

my small hosted farm mines more btc daily now then it did last June 2025

[MiningIntel]

That's a really good point. Difficulty growth definitely isn't a straight line, and your historical numbers make that clear. I like the idea of looking at multiple difficulty scenarios rather than assuming one fixed annual growth rate. Thanks for sharing the real-world data.

[FP91G]

That's a really good point. Difficulty growth definitely isn't a straight line, and your historical numbers make that clear. I like the idea of looking at multiple difficulty scenarios rather than assuming one fixed annual growth rate. Thanks for sharing the real-world data.

Increases in difficulty are usually offset by the price of Bitcoin. If the price doubles and the difficulty doubles, then with a simple calculation, your profit remains constant. But if the price halves and the difficulty doesn't, the miner enters a period of rut, in which only large companies survive.