pull down to refresh

Bitcoin’s Only Scaling Limitation Is Its Supply

Why No More Than 5-10 Million People Will Ever Be Able To Use Bitcoin Sovereignly
Bitcoin’s scaling debate is often framed as a bandwidth problem—how to fit more transactions into the blockchain—but this is a misdirection. The real limitation isn’t transactions per second, it’s who can afford to use the chain at all.
Scaling proposals, particularly those involving new opcodes, are disingenuous because they don’t actually allow more people to use Bitcoin on-chain. Instead, they simply create pooled transaction structures that shift security assumptions away from Bitcoin's trustless model. These designs fundamentally introduce custodianship, reliance on coordinators, and lower the barrier for Sybil attacks, making them indistinguishable from any other shitcoin.
At its core, Bitcoin’s scalability limitation is its supply—not block size, not opcodes, not transaction pooling. And because Bitcoin’s supply is permanently capped at 2.1 quadrillion satoshis, that means only 5-10 million people will ever be able to use it in a sovereign way.

1. Bitcoin’s True Limit: The 2.1 Quadrillion Satoshi Cap

Bitcoin is often described as having a 21 million BTC cap, but that’s just an abstraction. The real cap is 2.1 quadrillion satoshis—the smallest indivisible unit of Bitcoin.
  • Total supply: 2.1 quadrillion satoshis
  • Minimum viable amount per user: At least 20,000+ sats (enough to afford transaction fees over time).
  • Maximum possible sovereign users:
    2.1 × 10^15 sats ÷ 20,000 sats/person = 105 million users
    But this assumes perfectly equal wealth distribution, which does not exist.
In reality:
  • Whales and organizations hold a substantial percentage of all Bitcoin.
  • Millions of BTC are lost, locked up, or inactive.
  • Most people will never hold enough sats to transact meaningfully.
The result? The actual number of people who can afford to use Bitcoin in a sovereign way (without custodians or trusted third parties) is 5-10 million at most.

2. The Minimum Cost of On-Chain Transactions

Bitcoin isn’t free to use. Every transaction:
  • Takes ~200-300 bytes.
  • Requires at least 1 sat per byte in fees.
  • Cannot send less than 546 sats due to dust limits.
At just 1 sat per byte, a 200-byte transaction costs 200 sats.
If someone only owns 20,000 sats, a single transaction at 200 sats already eats up 1% of their total holdings. A few transactions, and they’re locked out of the network forever.
Even if someone is holding Bitcoin, if their balance is too low, they cannot afford to move it, which functionally excludes them from the system.

3. Scaling Proposals Are a Distraction

Proposals to “scale” Bitcoin using new opcodes and transaction pooling attempt to sidestep this reality. The issue isn’t how many transactions can fit into a block, it’s how many people can afford to go on-chain in the first place.
Scaling proposals are fundamentally dishonest because:
  • They don’t reduce the base cost of using Bitcoin.
  • They don’t increase the number of people who can afford to be sovereign.
  • They only bundle transactions together in a way that increases trust assumptions.
Pooling transactions together via new opcodes does not make Bitcoin more accessible. Instead, it shifts security into a coordinated system, requiring trusted operators who can go offline, harvest meta-data, collude to rug-pull users, or otherwise become centralized attack vectors.
If a pooling system anchors a group to the chain using one on-chain UTXO, what happens when a malicious actor, coordinator failure, or other complication needs to be resolved?
  • The users must take their disputes on-chain—but if they can’t afford on-chain fees, they have no recourse.
  • Attackers within the group can leverage Sybil attacks, overwhelming smaller users who cannot afford to challenge fraudulent transactions.
  • The “solution” becomes no different from a shitcoin—just another system making negative security trade-offs.
Bitcoin is what it is, you can accept security trade-offs for yourself, but that thing you accept is not Bitcoin, and those trade-offs are definitely not “scaling” Bitcoin.

4. Layer 2 (Lightning) Doesn’t Fix This

No one has spent more time building for Lightning than I, and I'll be the first to tell you that while moving transactions off-chain does a lot of great things, scaling to new users isn't one of them. Lightning still requires an on-chain transaction to open and close a channel.
If a user cannot afford an on-chain transaction, they cannot afford to enter Lightning in the first place. This means:
  • Lightning does not make Bitcoin cheaper for new users.
  • Lightning does not increase the number of sovereign users.
  • Lightning simply creates a second layer for those who can already afford on-chain fees.
The problem isn't TPS—it’s the cost of transacting at all.
At scale, Bitcoin’s model is not one of “everyone running their own node,” it’s one of a small minority having full sovereignty while the majority are priced into custodial services.

5. A Hard Fork for More Divisibility Won’t Happen

A hard fork to enable millisats (sub-satoshi units) would be a direct supply increase from 2.1 quadrillion satoshis to 2.1 quintillion millisats.
Bitcoin’s supply is hardcoded as an integer—it is not floating point, meaning there is no easy way to “add more decimal places” without actually changing the supply itself.
This would:
  • Require a hard fork that breaks most shell applications.
  • Violate Bitcoin’s fixed supply assumptions.
  • Be politically impossible.
Because a hard fork is unlikely to ever happen, Bitcoin will remain permanently constrained by its 2.1 quadrillion satoshi cap—ensuring that only a small number of people can ever use it directly.

6. The Inescapable Conclusion

Bitcoin’s only scalability limitation is its supply. Because:
  • A hard cap of 2.1 quadrillion sats exists.
  • On-chain transaction costs create a permanent user barrier.
  • True L2 solutions still require on-chain transactions to participate.
  • New opcodes and pooling mechanisms don’t increase the number of people who can use Bitcoin trustlessly
  • The only actual scaling solution, a hard-fork supply increase, is effectively impossible.
  • Denial of these facts is cope or shitcoining.

The result?

Bitcoin can only support 5-10 million sovereign users, ever.
Everyone else will either be priced out or forced into trust-based solutions—which isn’t really Bitcoin.
Every attempt to “scale” Bitcoin beyond this limit is just smuggling in trust, creating centralized points of failure that are indistinguishable from any other shitcoin.
The practical way forward is to keep trust localized to the family or community level, with tools like Lightning.Pub, and to make peace with this reality and love Bitcoin for what it is: Numéraire.
33 sats \ 3 replies \ @OT 12h
Sounds a bit defeatist. I think we can do better batching coordination or a quantum resistant taproot derivation path to get 10-20k TX into a block. That would allow for something like 100-200 million sovereign bitcoiners.
Cannot send less than 546 sats due to dust limits.
I think eventually the dust limit will need to be removed.
reply
10-20k TX into a block
The whole premise is that bandwidth isn't the problem.... Lightning already has infinite transactions in a block.
Defeatist
This is a good thing, it reaffirms Bitcoin is Numeraire and that scaling alarmists are complete morons
reply
0 sats \ 1 reply \ @OT 11h
Not sure if I like the idea of being part of a new elite. I also think it's too early to tell if it's going to be that way.
It is true that not many want to take responsibility for using bitcoin in a sovereign way. That could change when the need for it increases.
reply
You're slightly less elite now I was off by an order of magnitude, but still not enough to have us tx bound: #912705
reply
22 sats \ 1 reply \ @DarthCoin 18h
The practical way forward is to keep trust localized to the family or community level
I also have this "vision", that from now on, OG bitcoiners, knowledgeable and well resourced node runners, should start building local small community or familiar "LN banks", where the costs of opening/closing channels will be "shared" and help more normal users to use Bitcoin, but still with a limited trust.
In the past I even wrote a guide about how to build these "private LN banks" and what solutions are available.
I've tested your solution with Lightning.Pub too and I like it, is very easy to start and use. We should make more noise about it, even that is in an early stage, is still usable and only like that we can make it better, testing and using it.
reply
Someday they will remember Darth's words
Pub
New landing page is a step, trying to get a big wallet/pub co-release out the door soon... it's already the best at what it does but next step is to make people care
reply
Thanks for posting. Brings ideas back to reality.
reply
Really appreciate this discussion.
I guess this means we are ossified, now.
reply
reply
reply
  • Screwed up an order of magnitude somewhere along the way, should be closer to a billion on the high end which still does not change the tx pressure. There's only a billionish households and small businesses combined on the planet, and with Lightning they still won't be bandwidth limited as even today we can open a billion channels in just a single years worth of blocks.
reply
Wait I'm confused... if they can open a billion channels in a year... Then what's the problem? Open hundreds of millions of channels (there's tons of transaction fees right there) and then just...
use Lightning not touching on chain?
reply
Lightning inherently touches the chain, just having a channel open ties up sats
Who said it's a problem? Bandwidth is solved, ownership distribution is the limiting factor
reply
You are off by an order of magnitude. Even by two, considering that lightning already is in milisats, and is not custodial.
reply
Millisats aren't real, they're used for accounting not settling on chain
Also #912607
reply
0 sats \ 1 reply \ @bordalix 1h
1 sat/vbyte is a relay rule, not a consensus rule: a 0 fee tx is perfectly valid. Nodes don't relay tx with less than 1 sat/vbyte to prevent spam attacks, but they accept 0 fee txs in blocks.
This limitation was introduced when BTC was worth a couple of bucks, it can be lowered to 0.1 sats/vbyte (or even lower) while maintaining the original security assumptions.
So, a 138 vbytes tx would need to pay at least 14 sats to be relayed by the nodes.
0 sats \ 1 reply \ @mf 4h
How can the supply be a limitation when there is virtually infinite divisibility?
reply
There's no divisibility, read the post
reply
I don't believe that Bitcoin 'doesn't scale.' Try closing on a house (property) and paying 6% in closing fees.
On a house (for example at 100,000$) that's over 6k. And that's a cheap house in North America. On a 300k house that's ~18k just in closing fees.
The idea that paying to open a few Lightning Channels to send digital gold/digital capital around the internet is too expensive... is ridiculous.
If we got "just" say 7% of the world's population into using Bitcoin regularly (like less than 1 in 10) that's still 560 million people. That's ALOT
There are approximately 200 million on-chain transactions per year. Assuming that kind of on-boarding takes 10 years (at least) and say 1 in 3 self-custodies (which would be a higher % than today) that's...
186 million people using Bitcoin regularly in self-custody over 10-years time which I believe is not unreasonable... considering one lightning channel is enough for weeks or months and people can withdraw directly to a lightning channel to spend from.
reply
BTC doesn't scale.
What do you think stops a second Bitcoin chain from starting?
reply
Nothing, there's already countless of them... all the earliest shitcoins are Bitcoin forks.... but they're all worthless because you can't invent digital scarcity twice or duplicate Bitcoin's security.
reply
Why can't the same consensus rules with a mining algo change produce similar security?
reply
Will you be asking the same question over and over, expecting different answer?
reply
Security comes from the scale of the network, for which there is inherently one Schelling point, the rules are irrelevant
reply
Obviously many people use other chains.
Mostly the consensus rules of those chains are not even close to Bitcoin. But in a sustained high fee environment, I don't see why a second Bitcoin couldn't emerge.
reply
I don't understand this conversation. People want Bitcoin... not 'other chains'. Have you guys seen the eth/btc chart lately?
People also use dollars, dollars are arguably better currency.
Litecoin is arguably better currency than Bitcoin too... could go on forever with a million other shitcoins, stablecoins, blah blah blah.
Yet a second Bitcoin cannot emerge, because Bitcoin is not just a currency. Misunderstanding this fundamental thing is what that leads people into crypto more broadly.
we've argued about this elsewhere, but I still don't understand how you claim
"Requires at least 1 sat per byte in fees"
Isn't this just a node implementation / config detail re which transactions to relay, and not something. inherent to the protocol?
reply
A miner could mine a tx altruistically for no fee at all of course, or via some out-of-band coordinated lifetime fee plan... But you're still reliant on either someone else's altruism / something external and trusted to socialize the cost
Probably could tweak relaying to lower minimums, but that hasn't happened in part because going below the base units becomes indistinguishable from spam
reply
Probably could tweak relaying to lower minimums, but that hasn't happened in part because going below the base units becomes indistinguishable from spam
I disagree.
Min relay fee was introduced in 2013, when BTC was worth ~$100. A sat is now worth 1000 times more. If the value of a sat was sufficient to prevent spam in 2013, why can't we use a fraction of it today to prevent spam?
Nodes can use 0.1 sats/vbyte (or less) as min relay fee today. So, if a tx has 138 vbytes, it would need to pay at least 14 sats in fees (you need to round it up).
reply
Great, if you're certain then create a BIP to change it so we can tell the forkers to gfy
reply
reply
yes because we have mostly full blocks at 1 sat/vbyte.
but if we didn't there's no reason on principle that the relay cutoff level couldn't dip below till blocks filled up again. iiuc
hashrate would suffer.
but at the end, it's just a free market.
reply
Even if so, the larger point about consensus forks remain, they're pointless
reply