A blockchain is a distributed ledger that stores records in sequential blocks, each cryptographically linked to the one before it. No single party controls it. Every participant holds an identical copy of the full history. Once data is written, altering it requires rewriting every subsequent block across thousands of independent nodes simultaneously.
​

The name is literal: a chain of blocks. Each block contains a batch of validated transactions, a timestamp, and a cryptographic fingerprint of the previous block called a hash. That hash dependency is what binds the chain together. Change a single character in block 400, and the hash for block 401 no longer matches, invalidating block 401 too, and every block after it in turn.
​

Bitcoin popularized the design in January 2009. Satoshi Nakamoto's whitepaper described a peer-to-peer electronic cash system that could transfer value without a bank acting as intermediary. The core insight was that cryptographic linking plus distributed replication could replace institutional trust. A shared database where every participant independently verifies every transaction removes the need to trust any single keeper of that database.
​

Today the same architecture underlies thousands of networks: payment rails, smart contract platforms, tokenized assets, and stablecoin infrastructure. The foundational mechanics are the same whether you are looking at Bitcoin's simple UTXO ledger or Ethereum's account model running arbitrary code. Blocks, hashes, and distribution are the constant skeleton. What varies is what each network allows those blocks to contain.
​

For a closer look at how newer chains extend this architecture into layered execution environments, see what is a Layer 2 blockchain.
​

Consensus is the mechanism by which thousands of independent nodes agree on a single version of the ledger without a central authority deciding. Two dominant approaches have emerged: Proof of Work, which requires computational effort to earn the right to add a block, and Proof of Stake, which requires validators to lock up capital as collateral. Both make dishonest behavior economically painful.
​

In Proof of Work, miners race to solve a hash puzzle. The puzzle requires finding a number (called a nonce) such that when combined with the block's data and hashed, the output starts with a required number of zeros. This process consumes real electricity. The first miner to solve it broadcasts the block to the network. Other nodes verify the solution in milliseconds -- verifying is cheap, finding the answer is expensive. Bitcoin uses this mechanism. Its difficulty adjusts automatically every 2,016 blocks to target a 10-minute average block time regardless of how much mining power joins or leaves the network.
​

Proof of Stake replaces computational work with economic stake. Validators deposit (stake) the network's native token as collateral. The protocol selects validators to propose and attest to blocks, typically weighted by stake size. If a validator approves a fraudulent block, the protocol slashes a portion of their deposit. Ethereum transitioned from Proof of Work to Proof of Stake in September 2022, an event the community called The Merge. Validators must now stake a minimum of 32 ETH to participate directly.
​

Finality describes how certain a block's confirmation is. In Proof of Work, finality is probabilistic: a block buried under six additional blocks on Bitcoin is considered effectively final because the energy cost of replacing it exceeds any rational attacker's budget. Proof of Stake networks like Ethereum have explicit finality checkpoints: once two-thirds of validators attest to a checkpoint epoch, that history is cryptographically finalized and cannot be reverted without destroying at least one-third of all staked ETH.
​

Other consensus variants include Delegated Proof of Stake (used by older chains like EOS), Proof of History (Solana's time-stamping mechanism layered beneath its Proof of Stake tower), and Avalanche's probabilistic sampling consensus, which achieves fast finality by having validators repeatedly sample small random subsets of the validator set until the network converges. Each design makes tradeoffs between throughput, decentralization, and finality speed.
​

Learn more about how the trustless guarantee emerges from consensus design rather than from any single operator's honesty.
​

External reference: Ethereum's consensus mechanism documentation covers both PoW history and current PoS implementation in depth.
​

Immutability in a blockchain comes from hash chaining. Every block carries the hash of the block before it. A hash function is a one-way mathematical operation: it takes input of any size and produces a fixed-length output (256 bits in Bitcoin's SHA-256). Even a one-character input change produces a completely different output. There is no way to reverse the function to find what input produced a given hash.
​

When block N is written, it records the hash of block N-1. That means block N's own hash depends on N-1's contents. If you later modify a transaction inside block N-1, block N-1's hash changes. Now block N contains the wrong hash reference, making block N invalid. To fix block N you must recalculate it, which changes block N's hash, which breaks block N+1, and so on through every block that came after.
​

On a distributed network with thousands of nodes each holding the full chain, an attacker would need to redo all that work faster than the honest network is adding new blocks, and convince the majority of nodes to adopt the rewritten history. In Bitcoin's Proof of Work system this is known as a 51% attack: an attacker controlling more than half the network's total hashrate could theoretically rewrite recent history. For Bitcoin, that hashrate is in the hundreds of exahashes per second, making a sustained attack astronomically expensive. The immutability guarantee is economic, not purely mathematical, but the economics are so lopsided that the distinction rarely matters in practice.
​

Smart contract platforms add another layer. On Ethereum, once a contract is deployed to an address, its bytecode is stored onchain. The contract cannot be changed, only deprecated (if it was written with an upgrade mechanism) or replaced by a new contract at a different address. This is why auditing smart contract code before deployment matters: there is no patch cycle for onchain logic.
​

For users sending stablecoins or other tokenized assets, immutability is the foundation of settlement finality. When a USDC transfer confirms onchain, the record is permanent. No bank can reverse it, no processor can charge it back, and no single operator can retroactively alter the balance history. That property is what makes public blockchains useful for financial settlement rather than just record-keeping.
​

External reference: The original Bitcoin whitepaper at the Nakamoto Institute explains the hash-chaining mechanism and the probabilistic finality argument in sections 3 through 5.
​

A traditional database is controlled by a single operator who can read, write, update, and delete records. A blockchain is controlled by a distributed network of validators whose consensus rules make unauthorized changes computationally or economically impractical. The practical differences in control, transparency, and speed are significant and determine which tool fits which use case.
​

The choice between the two is not a competition: they solve different problems. If you are storing user profile data for a web app where you own all the records and need millisecond writes, a relational database is the correct choice. If you are settling a payment between parties who do not trust a common intermediary, recording an asset transfer that must be auditable by regulators and counterparties alike, or running logic that must execute identically no matter who calls it, a public blockchain's guarantees are worth the throughput tradeoff.
​

Hybrid architectures increasingly use both. An application might store large media files and searchable metadata in a database while anchoring proofs of that data onchain, combining the speed of a traditional database with the auditability of a blockchain.
​

For context on how rollups push blockchain throughput much closer to database speeds while keeping settlement guarantees, see what is a rollup.
​

External reference: Ethereum's intro documentation contrasts the Ethereum world computer model with traditional server architectures.
​

Not all blockchains are open to anyone. The three main categories differ in who can read the ledger, who can submit transactions, and who can validate blocks. Public chains are the most decentralized and the most widely used for financial applications. Private and permissioned variants trade some decentralization for performance and access control.
​

A public blockchain is open to any participant. Anyone can run a node, submit transactions, and on Proof of Stake networks, stake tokens to become a validator. Bitcoin and Ethereum are public. The transparency is total: every transaction is readable by any address or block explorer. Censorship resistance follows from this openness because no gatekeeper can block a transaction that meets the protocol's validity rules and pays sufficient fees.
​

A private blockchain is operated by a single organization. Hyperledger Fabric, used by some enterprise supply chain projects, is a common example. The operator controls who reads and writes. This model behaves more like a distributed database than a decentralized network: you are replicating data across several servers owned or trusted by one entity rather than achieving Byzantine fault tolerance across adversarial parties. The immutability guarantees are weaker because the operator can, in principle, modify the rules. The upside is much higher throughput and fine-grained access control.
​

A permissioned blockchain (sometimes called a consortium chain) sits between the two. A defined set of known organizations runs validator nodes. Any member of the consortium can validate, but outside parties cannot join without approval. R3 Corda and Quorum (now ConsenSys) are examples from financial services. Permissioned chains were popular in 2016 to 2019 enterprise blockchain pilots. Many organizations eventually found that a shared database with access controls served the same purpose with less complexity, while others migrated workloads to public chains as gas costs fell and Layer 2 networks matured.
​

For financial applications where stablecoins are involved, public chains are the dominant choice. USDC, USDT, and other major stablecoins exist primarily on public networks (Ethereum, Solana, Base, Avalanche, and others) because the liquidity, tooling, and counterparty reach on public chains exceed what any consortium can replicate.
​

External reference: Bitcoin.org's FAQ describes how Bitcoin's public, permissionless design was a deliberate choice rooted in censorship resistance.
​

Dozens of public blockchains exist, but four networks account for the large majority of stablecoin volume, DeFi activity, and developer tooling: Bitcoin, Ethereum, Solana, and Avalanche. Each made different engineering tradeoffs that shape what they are used for today.
​

Bitcoin launched in January 2009. Its block time is approximately 10 minutes and its throughput is roughly 7 transactions per second onchain. Those constraints are intentional: the design prioritizes security and decentralization over speed. Bitcoin's scripting language is intentionally limited, making complex smart contracts impractical on the base layer. Bitcoin's total supply is capped at 21 million BTC, with issuance halving approximately every four years. The most recent halving occurred in April 2024, reducing the block reward to 3.125 BTC.
​

Ethereum launched in July 2015. Its account model and Turing-complete EVM (Ethereum Virtual Machine) enabled arbitrary smart contracts for the first time at scale. ERC-20 tokens, DeFi protocols, NFT standards, and stablecoin infrastructure were all built on Ethereum's programmability. After The Merge in September 2022, Ethereum moved to Proof of Stake. Its base layer processes roughly 15 to 30 transactions per second, but the Layer 2 ecosystem (Arbitrum, Optimism, Base, zkSync) extends effective throughput by orders of magnitude while settling on Ethereum's security.
​

Solana launched in March 2020. It uses a combination of Proof of History (a cryptographic clock that sequences events before they are included in blocks) and Tower BFT consensus. Solana's theoretical throughput is 65,000 TPS, though real-world sustained throughput under load has been lower. Block times are approximately 400 milliseconds. Solana has experienced several network outages, most notably in September 2021, but has maintained higher uptime since 2022 through client software improvements. It is the dominant chain for high-frequency trading, consumer payments, and DeFi use cases where sub-second finality matters.
​

Avalanche launched in September 2020. Its primary innovation is a multi-chain architecture: the X-Chain handles asset transfers, the C-Chain (EVM-compatible) runs smart contracts, and the P-Chain coordinates validators. Avalanche's consensus achieves finality in under 2 seconds. The network also supports Subnets, which are custom blockchains that inherit Avalanche's validator set or use dedicated validators, enabling application-specific chains with shared security. AVAX is the native gas and staking token.
​

External reference: Ethereum's official history page documents every major upgrade from Frontier through The Merge and beyond.
​

A stablecoin is a token whose value is pegged to a reference asset, most commonly the US dollar. Stablecoins exist onchain as smart contracts that implement a token standard. On Ethereum and EVM-compatible chains, that standard is ERC-20. The token contract tracks balances, handles transfers, and enforces rules like minting authority. The issuer holds reserves offchain and mints or burns onchain tokens to maintain the peg.
​

USDC is the clearest example of how this works. Circle, USDC's issuer, holds dollars and short-term US Treasury bills in regulated accounts. When an authorized participant deposits fiat, Circle's minting contract creates an equivalent number of USDC tokens onchain. When the participant redeems, the tokens are burned and fiat is released. The onchain record of every USDC transfer is publicly auditable. The trust you extend is to Circle's reserve attestations and to the ERC-20 contract's code, not to any private database.
​

For a detailed explanation of how the ERC-20 standard governs token behavior (approvals, allowances, transfer events), see what is an ERC-20 token.
​

Moving USDC across chains historically required bridging protocols that locked tokens on one chain and minted wrapped representations on another. Circle's Cross-Chain Transfer Protocol (CCTP) changed this. CCTP uses a burn-and-mint model: USDC is burned on the source chain and natively minted by Circle on the destination chain. No wrapped token exists. The receiving chain gets canonical USDC backed by the same Circle reserves, not a bridge-wrapped proxy. CCTP is live on Ethereum, Base, Avalanche, Solana, and a growing list of networks.
​

USDT (Tether) follows a similar model but operates under different reserve disclosures and is issued by Tether Limited. DAI is a decentralized stablecoin issued by the MakerDAO protocol: it is minted by depositing collateral (ETH and other assets) into onchain vaults rather than relying on a centralized issuer holding fiat. Each model represents a different point on the spectrum between issuer-trust and protocol-trust.
​

For how stablecoin automation platforms layer yield, routing, and compliance logic on top of these onchain primitives, see the best stablecoin automation platforms.
​

External reference: Ethereum's stablecoins overview explains the differences between fiat-backed, crypto-backed, and algorithmic stablecoin designs.
​

No. A blockchain is the underlying ledger technology. A cryptocurrency is a native token that an application or protocol issues on top of it. Bitcoin is both a blockchain and a cryptocurrency with the same name. Ethereum is a blockchain; ETH is its native token. Many blockchains host hundreds of tokens that are not the chain's own cryptocurrency.
​

Not on a public blockchain. Confirmed blocks are permanent. While full nodes can prune older transaction data to save disk space, the block headers and merkle roots remain, preserving cryptographic proof of history. There is no admin account that can erase a transaction once it reaches finality. This is the core distinction from any centrally managed database.
​

It depends entirely on the chain. A Bitcoin transaction typically reaches practical finality after about 60 minutes (6 confirmations). Ethereum base-layer finality takes roughly 12 to 15 minutes to reach a finalized checkpoint. Solana achieves sub-second block inclusion. Layer 2 networks on Ethereum can confirm transactions in 1 to 2 seconds while inheriting Ethereum's settlement guarantees after a delay.
​

A smart contract is code deployed onchain that executes automatically when predefined conditions are met. It lives at a blockchain address just like a wallet, but instead of being controlled by a private key, it runs logic written by its developer. Token contracts, lending protocols, automated market makers, and stablecoin minting systems are all implemented as smart contracts. Once deployed, the code cannot be changed without a deliberate upgrade mechanism.
​

Blockchain-based payments settle in seconds to minutes rather than 1 to 3 business days, operate 24 hours a day without banking hours, and are accessible from any wallet without a bank account. For cross-border transfers, they bypass correspondent banking chains that add fees and delays. Stablecoins combine dollar-denominated value with onchain settlement speed, making them a practical payment primitive for both consumer and business use cases.
​

Related reading