Bitcoin History • Mining • Decentralization

Keeping Bitcoin Decentralized

What the failure of BIP-110 teaches us about consensus, why concentrated mining pools deserve scrutiny, why changing Bitcoin from SHA-256 to CPU/GPU mining is probably the wrong cure, and what ordinary Bitcoiners can actually do to make the network harder for governments, corporations, financial institutions and large pool operators to capture.

21MMaximum bitcoin supply
SHA-256dBitcoin proof of work
2016Blocks per difficulty period
~10 minTarget block interval
Decentralized Bitcoin mining networkMany independent nodes and miners surround a Bitcoin symbol instead of connecting through one central controller.NODEMINERNODEMINERNODEMINERNODEMINER
Central idea
Bitcoin is strongest when no single group controls all the important layers. Decentralization means distributing validation, mining hardware, block-template construction, custody, software development, network access and economic ownership—not merely counting how many mining-pool logos appear on a chart.

BIP-110: a decentralization lesson written into Bitcoin history

BIP-110, titled Reduced Data Temporary Softfork, proposed temporary consensus restrictions intended to limit large arbitrary-data fields in Bitcoin transactions. Its supporters argued that Bitcoin should remain primarily a monetary network and that large data inscriptions impose costs on node operators and compete with financial transactions for block space.

The concern itself was not irrational. Bitcoiners can legitimately disagree about inscriptions, arbitrary data, relay policy and how much burden node operators should be expected to carry. The crucial historical question was not whether the concern was sincere—it was whether enough of the Bitcoin economy was willing to adopt the new consensus rule.

December 2025: the BIP's deployment window began, using version bit 4 and a modified activation mechanism.
Threshold: BIP-110 specified a 55% signaling threshold—1,109 of 2,016 blocks—for early lock-in.
Block 961,632: its mandatory-signaling period began. Enforcing nodes were programmed to reject blocks in the window that did not signal bit 4.
Chain split: reported miner signaling was only about 2.53%. Instead of forcing the larger Bitcoin network to comply, the enforcing nodes separated onto a very small minority chain.
Stalled mining: reporting documented the minority chain producing only two blocks while the main chain rapidly moved dozens of blocks ahead.
August 9, 2026: the official BIP repository changed BIP-110's status to Closed, with the changelog stating that it was closed following a chain split with stalled mining.
Important: “BIP-110 failed” does not mean “Bitcoin must accept every use forever.”
Node operators may still choose local mempool and relay policies. Miners may choose which valid transactions they put in their own candidate blocks. What failed was the attempt to make a disputed transaction restriction a consensus rule without broad enough support.

Policy is not the same thing as consensus

This distinction is one of the most important lessons from BIP-110. A node can say, “I do not want to relay this transaction,” while still accepting a valid block containing that transaction after a miner includes it. That is policy. A consensus rule says, “A block containing this transaction is invalid.” Consensus disagreements can split the chain.

DecisionLocal policyConsensus rule
Relay a transaction?Node operator can chooseNot normally required for block validity
Keep transaction in mempool?Node operator can chooseDoes not by itself change Bitcoin rules
Put transaction in a candidate block?Miner/template builder can chooseBlock still must satisfy consensus
Reject an otherwise-valid block?No longer merely policyThis changes the accepted consensus set

A UASF is not automatically wrong. Bitcoin history includes successful user pressure, most famously around SegWit. But BIP-110 demonstrated that the phrase “users decide” does not mean a small subset of users can unilaterally command miners, exchanges, businesses and other users. Bitcoin consensus ultimately emerges from voluntary coordination among economically relevant participants enforcing compatible rules.

The mining-pool problem: concentrated coordination

Bitcoin's physical hashrate is enormous and spread across many facilities and owners, but the public face of block production is concentrated in a relatively small number of mining pools. That matters because traditional pools often construct the candidate block template and tell participating ASICs what work to perform.

At the time this page was prepared on August 17, 2026, Hashrate Index's current pool table showed approximately:

Foundry USA
23.8%
AntPool
22.81%
F2Pool
15.34%
SpiderPool
8.85%
ViaBTC
8.55%

Snapshot only. Pool shares fluctuate continuously. The five figures above total about 79.35% of the observed window, illustrating concentration at the coordination layer—not necessarily common ownership of 79.35% of the world's ASIC hardware.

A mining pool is not necessarily one giant miner

This is frequently misunderstood. If a pool has 20% of reported hashrate, the pool company does not necessarily own 20% of all Bitcoin mining machines. Independent companies, farms and individuals can point their ASICs at the same pool because pooling reduces payout variance.

But concentration still matters. If the pool builds the template, the operator can influence which transactions its participants attempt to confirm. A few pool operators can therefore become attractive pressure points for governments, regulators, sanctions regimes, banks or commercial partners that want certain transactions filtered.

🧱

Template concentration

If a few pools choose most block templates, transaction-selection authority is more centralized than the ASIC ownership chart suggests.

🎯

Regulatory chokepoints

Large identifiable pool companies can be easier to regulate, sue, sanction or pressure than thousands of independent home miners.

💰

Payout dependence

Miners join pools for predictable revenue. That economic need can keep hashrate concentrated even when miners philosophically prefer independence.

The good news: pool concentration is unusually reversible.
An ASIC can often change pools in minutes. Hashrate is not permanently married to a pool. If miners retain the freedom and technical ability to switch, a pool that censors, changes terms or becomes politically captured can lose hashrate quickly.

What 51% of hashrate can—and cannot—do

A majority of hashrate is dangerous because it can potentially reorganize recent history, censor transactions by refusing to include them, or attempt double-spend strategies. But hashrate does not grant the power to create arbitrary bitcoins, spend coins without keys, rewrite the 21-million limit by itself, or force validating nodes to accept blocks that violate the nodes' consensus rules.

That is why Bitcoin's decentralization is layered. Miners propose blocks. Full nodes independently validate them. Wallet owners control keys. Markets decide what they value. Developers publish software, but users decide what to run.

Governments, Wall Street and corporations: influence is real, control is different

Government pressure

Governments can exert substantial influence around Bitcoin's edges. They can regulate exchanges, require identity verification, pressure public mining companies, control access to electrical grids, restrict imports of mining equipment, enforce sanctions, tax mining income and regulate custodians. A mining industry dominated by a handful of giant, publicly visible facilities is easier to pressure than a geographically dispersed mixture of industrial, commercial and home miners.

But a government cannot simply issue a regulation that causes independent full nodes around the world to accept an invalid transaction or an inflationary block. To gain that kind of control, it would need much broader technical, economic and social coordination.

Wall Street and institutional investors

Institutional investment changes Bitcoin's market structure. ETFs, large custodians, public mining firms, derivatives markets and corporate treasuries can influence price discovery, liquidity and public perception. Concentrated custodial holdings can create operational or political chokepoints.

However, owning a billion dollars of bitcoin does not provide a billion dollars' worth of protocol votes. Bitcoin is not proof-of-stake. A large investment company cannot change the 21-million limit merely because it owns a large quantity of BTC.

The defensive principle: the more bitcoin that remains in genuine self-custody, the more users run validating nodes, and the more miners construct their own templates, the harder it becomes for financial intermediaries to turn ownership concentration into protocol control.

Corporate mining

Large professional miners are not automatically enemies of Bitcoin. They contribute enormous proof-of-work security and can develop efficient energy infrastructure. The risk appears when mining hardware ownership, pool selection, template construction, firmware, energy supply and custody all become concentrated in the same small set of organizations.

Bitcoin does not need to eliminate profitable mining. Bitcoin needs profitable mining to remain contestable: new entrants must be able to join, miners must be able to switch pools, independent operators must be able to validate the chain, and users must be able to withdraw to wallets they control.

Should Bitcoin abandon SHA-256 ASICs for CPU/GPU mining?

At first glance, switching Bitcoin to a CPU-friendly or GPU-friendly proof-of-work algorithm seems like a direct way to destroy industrial ASIC dominance. Everyone already owns a CPU; millions own GPUs. Why not make ordinary computers miners again?

Because the cure would introduce new and potentially larger risks.

IssueKeep SHA-256 ASIC PoWFork to CPU/GPU-oriented PoW
Existing security capitalPreserves global SHA-256 mining infrastructureStrands existing Bitcoin ASIC investment on the fork
Hard-fork riskNo algorithmic chain split requiredVery likely creates competing chains if old miners continue SHA-256
Hardware purposeASICs are specialized for SHA-256CPUs/GPUs are general-purpose and rentable
Cloud concentrationIndustrial miners can centralizeLarge cloud providers and data centers own huge general-purpose fleets
Botnet incentiveConsumer computers cannot efficiently mine SHA-256 BTCCPU-friendly PoW can create incentives for malware mining
Long-term specializationSpecialization is explicitASIC resistance may reduce, not permanently eliminate, specialization
Consensus stabilityPreserves Bitcoin's long-established PoW ruleChanges one of Bitcoin's deepest consensus assumptions

Why SHA-256 ASICs can actually help Bitcoin

An Antminer has very little alternative economic purpose. A Bitcoin SHA-256 ASIC represents sunk capital dedicated to proof of work. That specialization makes the security resource less interchangeable with cloud-computing workloads, AI workloads, video rendering or general corporate computing.

A CPU/GPU network can be more accessible at the hardware level, but its computing power is also more general-purpose and may be easier to rent or redirect. Large cloud providers, universities, data centers and compromised computer networks already possess enormous pools of general-purpose computation.

RandomX is a serious example of a CPU-oriented PoW design. Its own documentation states that it is optimized for general-purpose CPUs and uses random code execution plus memory-hard techniques to minimize the efficiency advantage of specialized hardware. That is a valid design goal for Monero. It does not follow that Bitcoin should abandon its existing SHA-256 security ecosystem to pursue the same tradeoff.

“ASIC resistant” is not the same as “centralization resistant”

Even if an algorithm prevents a large ASIC advantage, economics still rewards cheap electricity, low cooling costs, access to capital, bulk hardware purchasing and professional operations. GPU farms can become industrial. CPU farms can become industrial. Cloud computing is already industrial.

And miners would still join pools because solo-mining variance does not disappear merely because the hashing device is a CPU. Millions of CPUs can still point to three pools.

Changing the chip does not automatically change who constructs the block.
If ten million CPU miners all accept templates from three centralized pools, Bitcoin's transaction-selection layer could remain highly concentrated. The more direct solution is to decentralize block-template construction.

Why keeping SHA-256 is the better course

Bitcoin's SHA-256 proof of work has accumulated a massive ecosystem of hardware, firmware, energy contracts, engineering expertise, repair infrastructure and operational knowledge. That installed capital raises the cost of attacking the existing network. Throwing it away to reset mining hardware would exchange a known centralization challenge for a new and unpredictable security environment.

The better strategy is to make SHA-256 mining itself more distributed and sovereign: more owners, more locations, more energy sources, more independent nodes, more miner-created templates and easier pool switching.

What can actually make Bitcoin more decentralized?

🖥️

1. Run a full node

Your own node verifies Bitcoin's rules. Do not outsource consensus verification to an exchange, explorer or pool when you can verify locally.

⛏️

2. Expand home mining

Home, farm, workshop, hydro, solar and heat-reuse mining spreads physical hashrate across many owners and jurisdictions.

🔑

3. Self-custody

Decentralized money held entirely by centralized custodians recreates a chokepoint. Withdraw to wallets whose keys you control.

🧩

4. Build your own templates

Use protocols that let the individual miner—not merely the pool—choose valid transactions for candidate blocks.

🔀

5. Switch pools when needed

Miners should treat a pool as replaceable infrastructure. Censorship or unacceptable policy should cause hashrate to leave.

🛠️

6. Preserve software choice

Core, Knots and compatible tooling should be evaluated on merits. Diversity is useful when consensus compatibility is respected.

DATUM: turn the pool into a payout coordinator

OCEAN's DATUM—Decentralized Alternative Templates for Universal Mining—is designed to let miners create block templates from their own Bitcoin node. The miner can still participate in a supported pool for payout smoothing, while transaction selection moves back toward the miner.

Your Bitcoin node
validates chain and maintains your mempool
DATUM gateway
creates/distributes your mining work
Your SHA-256 ASIC
performs proof of work

The crucial idea is that the pool can become more like an accountant than a block editor. Pooling can remain useful for low-variance payments without requiring one corporation to dictate every transaction its members attempt to confirm.

Stratum V2 Job Declaration

Stratum V2 attacks the same centralization problem from the mining-protocol layer. Its Job Declaration Protocol is explicitly designed to prevent pools from unilaterally imposing work on miners. In the design, pools can focus on accounting for shares and distributing rewards while miners declare custom work. The specification also provides fallback behavior if a pool rejects valid custom work.

OLD POOL MODEL Pool operator ↓ selects transactions ↓ constructs template Thousands of miners ↓ provide hashpower MORE DECENTRALIZED MODEL Independent miner's full node ↓ selects valid transactions Miner-side template builder ↓ declares work ASIC hashes the miner's template ↓ Pool coordinates shares and payouts

Solo mining still has a role

True solo mining gives the individual miner maximum independence. The miner runs the node, constructs the template, performs the work and—if lucky enough to find a valid block—receives the full subsidy and transaction fees.

The disadvantage is brutal payout variance. A single modern home ASIC represents only a tiny fraction of global hashrate. Solo mining is therefore excellent for sovereignty and experimentation but can go extraordinarily long periods without any reward. Pooling exists because that financial variance is real.

That is why miner-selected templates plus pooled payouts may be one of the most practical decentralization advances available today.

Make pool failure survivable

A healthy mining setup should avoid a single point of failure. Miners can configure alternative pools, keep local node infrastructure maintained, understand how to change Stratum endpoints, and periodically verify that the pool's policies still match their own expectations.

✓ Independent validation: run your own Bitcoin full node.
✓ Independent custody: hold keys outside exchanges and ETF structures.
✓ Independent templates: use DATUM or Stratum V2-style miner construction where practical.
✓ Physical dispersion: encourage small and medium miners alongside industrial miners.
✓ Easy exit: keep the ability to leave a pool quickly.
✓ Verify, don't trust: watch your node, template source, payout rules and software updates.

What BIP-110 should teach the next generation of Bitcoiners

Lesson 1: Consensus is hard to change—and that is a feature. A monetary system intended to last generations should not change fundamental rules merely because one developer, corporation, government, mining pool or online faction demands it.

Lesson 2: Node operators matter, but they cannot ignore economic reality. A node can enforce any rule its owner chooses. If almost nobody else accepts that rule, the node can isolate itself onto a minority chain. Sovereignty includes accepting the consequences of your own rules.

Lesson 3: Miners are powerful but not kings. Miners choose and order valid transactions and provide proof of work. They cannot force independently validating nodes to accept invalid blocks.

Lesson 4: Users are powerful when they take custody. Bitcoin held entirely through custodians gives those custodians leverage. Self-custody turns economic ownership into direct control of keys.

Lesson 5: Pool decentralization should focus on template authority. A pool with 25% of hashrate is much less threatening if thousands of participating miners independently construct 25% worth of candidate blocks.

Lesson 6: Do not throw away SHA-256 merely to reset the hardware market. Bitcoin's specialized proof-of-work infrastructure is part of its security. Decentralize ownership and decision-making around that infrastructure instead.

Lesson 7: Freedom requires exit. The ability to change software, move an ASIC to another pool, broadcast through another peer, withdraw from a custodian, move jurisdictions, or run your own node is more important than trying to design a system in which no powerful institution ever exists.

Bitcoin does not stay decentralized because powerful organizations promise to behave. It stays decentralized when ordinary participants retain the technical ability to verify, hold keys, build blocks, mine, route around failures and leave intermediaries.

A practical decentralization blueprint

For an individual Bitcoiner

Learn self-custody, back up recovery material securely, run a full node if practical, use your own node with your wallet, avoid leaving long-term holdings on exchanges, and understand what software rules your node is enforcing.

For a home miner

Keep SHA-256 hardware productive. Point it toward smaller or decentralization-focused pools when economically reasonable. Experiment with DATUM, Stratum V2 or true solo mining. Keep a fallback pool configured so a local server failure does not leave the ASIC idle.

For a mining company

Diversify pool relationships, support miner-created templates, avoid building dependence on one custodian or one jurisdiction, and make censorship policy transparent. Industrial miners can strengthen Bitcoin decentralization if they refuse to become captive hashpower.

For wallet and node developers

Make self-hosting, self-custody and miner-directed block construction easier. Good decentralization technology is technology ordinary people can actually operate.

For the Bitcoin community

Watch concentration without confusing size with guilt. Large organizations are not automatically hostile; small organizations are not automatically trustworthy. Judge systems by whether users can verify them and exit them.

The objective is not “no corporations.”
The objective is a Bitcoin network where no corporation, government, bank, ETF sponsor, mining pool, software team or wealthy investor can become indispensable.

Conclusion: decentralization is something Bitcoiners do

BIP-110 became an important 2026 case study because it showed both sides of Bitcoin sovereignty. Node operators were free to enforce a controversial rule. The rest of the network was free not to follow them. When the mandatory-signaling fork attracted insufficient mining support, the minority chain stalled and the proposal was subsequently marked Closed.

The mining-pool concentration problem is more enduring. A small number of pools currently coordinate a large majority of observed hashrate, which creates legitimate concerns about transaction censorship, regulation and infrastructure dependence. But replacing SHA-256 with CPU or GPU proof of work would not automatically solve that problem. It could destroy existing security capital, create a contentious hard fork and simply replace ASIC concentration with cloud, GPU-farm or general-purpose computing concentration.

The more promising route is evolutionary rather than destructive: keep SHA-256, spread ASIC ownership, expand home mining, run independent nodes, move block-template authority to miners, use protocols such as DATUM and Stratum V2, maintain the ability to switch pools, and keep bitcoin in self-custody.

Bitcoin's defense has never been that powerful actors disappear. Its defense is that the system gives independent participants enough tools to route around them.

← Return to BTC.TedLee.ca — Bitcoin Education

Important disclaimer

Educational and historical material only. This page is not financial, investment, legal, tax, cybersecurity or mining-business advice. Bitcoin mining can lose money. Bitcoin prices are highly volatile. Running experimental node or mining software can create operational risks, including following an unintended chain or losing mining revenue.

Pool market-share statistics change continuously. Verify current data before relying on any figures. Software, consensus proposals and mining protocols also evolve; check primary sources and release documentation before changing a production node or miner.

No statement on this page should be interpreted as an accusation that a named company, pool, government, financial institution or developer is acting maliciously. The discussion concerns structural concentration risk and how decentralized systems can reduce dependence on trusted intermediaries.

Sources & further reading

The most important claims on this page are grounded in primary protocol documentation and current mining data, with contemporary reporting used for the BIP-110 chain-split history.

  1. Bitcoin BIPs — BIP-110: Reduced Data Temporary Softfork. Primary specification. The repository shows status Closed; its August 9, 2026 changelog says it was marked closed following a chain split with stalled mining.
  2. CoinDesk — Controversial Bitcoin fork BIP-110 mines two blocks, then stops. Contemporary reporting on the minority chain and reported miner support.
  3. Hashrate Index — Bitcoin Mining Pool Data. Current and historical pool share data. Figures on this page were captured August 17, 2026 and will change.
  4. Stratum V2 — Job Declaration Protocol Specification. Primary specification describing custom work, miner-side job declaration and pool share/reward coordination.
  5. OCEAN — DATUM Setup Guide. Describes creating block templates with the miner's own Bitcoin node and mining via DATUM-supported pools or solo.
  6. OCEAN DATUM Gateway source repository. Open-source implementation of miner-side template creation and block submission.
  7. Bitcoin Developer Reference — Block Chain. Technical reference for Bitcoin's 80-byte block header and SHA-256d proof-of-work construction.
  8. RandomX official source repository. Primary documentation for a CPU-optimized, memory-hard proof-of-work design intended to minimize specialized-hardware advantage.
  9. BTC.TedLee.ca — Bitcoin Education. Main educational site.

Last substantive update: August 17, 2026. Because mining-pool shares and software status change rapidly, time-sensitive values should be independently rechecked.

↑ Back to top