Blockchain Almanac | Ethereum architecture guide
Navigation: How to use the Almanac · Previous ← Bitcoin architecture guide
Contents of this article
- Account model vs UTXO
- Proof of Stake and validator workflow
- Smart contracts and the EVM stack
- Gas, fees, and MEV
- Layer-2 rollups and data availability
- Upgrade roadmap from Shanghai to Danksharding
- What it means for traders
- Apply in trading with Veles
- Related materials
Account model vs UTXO
Ethereum abandons Bitcoin’s UTXO structure in favor of an account model: every address stores its balance and nonce, while smart contracts maintain their own state. This simplifies contract interactions (no need to aggregate “coins”), but introduces global state changes that every node must track.
For developers it means composability — DeFi protocols can call each other inside one transaction. For traders it explains why Ethereum DEXs easily daisy-chain swaps, but congestion affects the entire network because all accounts compete for the same block space.
Proof of Stake and validator workflow
Since “The Merge,” Ethereum relies on Proof of Stake. Validators stake 32 ETH, join an activation queue, and then propose or attest to blocks in 12-second slots. If they miss duties or attempt to finalize conflicting chains, they lose a portion of their stake (slashing).
The consensus layer (formerly Beacon Chain) rotates committees to keep leader selection unpredictable. Finality arrives after two epochs (~12.8 minutes), which is faster than Bitcoin’s probabilistic finality yet still enough time for traders to react to macro news before a transaction becomes irreversible.
Smart contracts and the EVM stack
The Ethereum Virtual Machine executes bytecode produced by Solidity, Vyper, or other high-level languages. Each operation has a gas cost that reflects CPU, memory, and storage usage. Storage writes are the most expensive, so protocols rely on mappings and events instead of frequently rewriting entire arrays.
Execution is deterministic: if a transaction reverts, all state changes roll back but the sender still pays for the gas consumed up to the failure point. That is why advanced bots simulate their transactions before submission and adjust slippage or calldata accordingly.
Gas, fees, and MEV
Since EIP-1559 the base fee is burned while users add a priority tip for validators. During volatile markets the base fee can spike above 200 gwei, making small trades unprofitable. On top of that, Maximal Extractable Value (MEV) bots reorder or sandwich transactions to capture arbitrage.
Strategies on Ethereum must budget for these dynamics: use limit orders or on-chain TWAP tools when gas is low, and prefer Layer-2 deployments for scalping. Monitoring mempool congestion is as important as the entry conditions you code into your bots — RSI, Bollinger Bands, volatility filters — because fees can invalidate a signal in seconds.
Layer-2 rollups and data availability
Optimistic rollups (Arbitrum, Optimism) batch thousands of transactions, publish compressed data to Ethereum, and allow one-week fraud proofs. ZK-rollups (zkSync, Starknet, Linea) submit validity proofs immediately, offering faster withdrawals at the cost of more complex cryptography.
Data availability remains on L1, so gas savings depend on calldata compression (e.g., EIP-4844 blobs). Traders gain cheaper execution without sacrificing the security of the base layer, but must mind liquidity fragmentation between L1 and each rollup.
Upgrade roadmap from Shanghai to Danksharding
Shanghai/Capella enabled staked-ETH withdrawals, increasing validator churn but also attracting institutional capital. The next milestones — EIP-4844 (proto-danksharding) and full Danksharding — will introduce blob-carrying transactions and spread data across multiple shards, slashing Layer-2 fees.
Each release tightens the relationship between execution, consensus, and data availability layers. Understanding timelines helps traders anticipate when gas-sensitive strategies (grid bots, DCA) become viable directly on Ethereum again instead of only on rollups.
What it means for traders
- Account model: Unlimited composability, but global state contention can spike gas in seconds. Always keep a reserve for unexpected fee jumps.
- Proof of Stake: Faster finality reduces counterparty risk when moving collateral between protocols, yet validator queues can delay large unstaking operations.
- EVM and gas: Storage-heavy contracts (NFT mints, airdrops) often crowd out regular swaps; schedule rebalancing outside of hype windows.
- Layer-2: Re-deploy bots across Arbitrum, Optimism, or Base to preserve edge while keeping the mainnet as a settlement rail.
Apply in trading with Veles
- BYBIT ETH BOMBERMAN 1.1 — mirrors the Bomberman setup where BOP and Mean Reversion Channel govern entries and exits, while the Donchian channel is used for entry signals on breakout boundaries.
- Veles Daily — cycles and flagships — weekly macro prompts to decide whether bots stay on mainnet or migrate to rollups before major news.
Related materials
What comes next
- How blockchains work — revisit consensus basics to contrast account vs UTXO models.
- Blockchain layers guide — map Layer-0/1/2 responsibilities before configuring bots on rollups.
- Bitcoin architecture guide — compare PoW finality with Ethereum’s PoS pipeline.
Glossary
- Account model — state accounting where each address stores balance and nonce instead of aggregating UTXO.
- Proof of Stake — consensus in which validators lock 32 ETH to propose and attest blocks.
- EVM — Ethereum Virtual Machine executing smart-contract bytecode.
- MEV — maximal extractable value captured by reordering transactions inside a block.
- Proto-danksharding — EIP-4844 upgrade introducing blob-carrying transactions for cheaper Layer-2 data.
- Full trader glossary