The blockchain trilemma explained

What is the blockchain trilemma?

The definitions most writers use come from a 2021 essay by Vitalik Buterin, which the Bank for International Settlements (BIS) cites when it discusses the idea. Buterin defines the three properties precisely:

• Scalability: the chain can process more transactions than a single regular node — think of a consumer laptop — can verify.
• Decentralisation: the chain can run without depending on trust in a small group of large actors.
• Security: the chain can resist a large share of its participants trying to attack it.

The trilemma is the claim that simple designs get at most two of these. The BIS states it bluntly in its 2022 Annual Economic Report: permissionless blockchains can achieve only two of the three. Notice that it is an engineering rule of thumb, not a mathematical theorem — which is why people keep trying to beat it.

Why can't a blockchain just process more transactions?

On a public chain, every full node re-checks every transaction. That is what lets ordinary users verify the ledger without trusting anyone. It also means the chain can only go as fast as the nodes you expect people to run.

Turn the dial up — bigger blocks, faster blocks — and each node needs more processing power, memory and bandwidth. The Lightning Network whitepaper put a number on this for Bitcoin in 2016: matching Visa's holiday peak of 47,000 transactions per second on-chain would need blocks of nearly 8 GB every ten minutes, something only a few large operators could handle.

Keep the dial down and block space stays scarce. The BIS explains that transactions offering higher fees are more likely to be included, so in busy periods users bid against each other. Its 2022 bulletin found Ethereum fees averaging around $1 in modest traffic could average more than 75 times higher on days of heavy congestion.

How do real blockchains choose their trade-offs?

Buterin's essay sorts traditional designs into three families, each giving up one corner of the triangle. The table follows his grouping; the examples in the first column are his.

The BIS applies similar logic to newer layer 1s, arguing that chains such as Avalanche and Solana gain capacity at the price of greater centralisation and weaker security. Supporters of those chains weigh it differently; see What is Solana? for how one such design works.

How does node hardware reveal the trade-off?

The clearest evidence is in each project's own documentation of what it takes to participate. The roles are not identical — a Bitcoin full node, an Ethereum full node and a staked Solana validator do different jobs — but the gap shows how much each design asks of its operators.

Buterin's argument is that users should not have to trust a set of nodes they could not join with a consumer laptop. By that yardstick, hardware requirements are not a technical footnote; they decide who gets to check the rules.

Can layer 2s or sharding get around the trilemma?

That is the bet Ethereum has made. ethereum.org explains that its scaling effort now centres on layer 2: rollups execute transactions elsewhere and post the data back, so the base chain stays light enough for ordinary nodes while total capacity grows. Bitcoin's Lightning Network applies a similar idea through payment channels.

Critics see costs elsewhere. The BIS argues that moving activity to layer 2s and rival chains fragments users and liquidity, and that bridges between chains often rely on a small number of validators — a weak point exploited in several high-profile hacks. Whether layer 2 designs truly escape the trilemma or just relocate its trade-offs is still debated.

What mistakes do people make when talking about the trilemma?

• Comparing raw transactions per second. A high number means little without asking who can verify it and at what hardware cost.
• Treating the trilemma as a law of physics. It is a widely cited rule of thumb; researchers are actively trying to weaken it.
• Counting validators instead of independence. Many nodes controlled by a few operators can be less decentralised than the headline count suggests — Buterin's test is whether you depend on a small group of large actors.
• Forgetting the bridge. A fast chain linked to others by a lightly secured bridge is only as secure as that bridge.