What Are Gas Fees in Crypto? A Complete Guide to Ethereum Network Costs

Rollups are Layer 2 scaling solutions that address the blockchain scalability trilemma by processing transactions off-chain while anchoring security to the main Ethereum chain. They achieve approximately 400x scalability improvement through transaction batching and data compression techniques like simplified gas fees and aggregated BLS signatures. There are two main types: Optimistic Rollups (e.g., Arbitrum, Optimism) which assume transactions are valid and allow a 7-day dispute period for fraud detection, and ZK-Rollups (e.g., zkSync, StarkNet) which use cryptographic zero-knowledge proofs for immediate transaction verification. Rollups are EVM-compatible, allowing developers to port existing Ethereum applications with minimal changes, and they significantly reduce transaction costs from 45,000 gas on mainnet to approximately 300 gas on rollups.

ZK (Zero-Knowledge) Rollups will lower transaction fees by another 40 to 100x compared to current layer 2 solutions. This is possible because ZK rollups only need to provide validity proofs to finalize transactions on Ethereum, instead of posting all their transaction data on chain like optimistic rollups. The high cost issue of optimistic layer 2 rollups comes from the data storage part needed to run the layer 2. By reducing the amount of data stored on the Ethereum layer 1, ZK rollups solve the transaction cost issue.

Layer 2 solutions scale Ethereum by processing transactions off-chain while maintaining security through the mainnet. Optimistic rollups assume transactions are valid unless proven otherwise, allowing near-instant deposits but requiring 7-day withdrawal periods for challenge windows. ZK-rollups use zero-knowledge proofs where verification time is independent of transaction complexity, though proof creation is computationally intensive. Leading solutions like zkSync promise 20,000 transactions per second. The 1inch Liquidity Protocol generates 2-3 times more rewards than standard DEX pools by capturing more trading volume, with impermanent loss occurring when tokens in pools change value relative to each other.

Layer one blockchains are inherently unscalable due to the scalability trilemma, forcing developers to choose between scalability, security, and decentralization. Those claiming scalability typically sacrifice decentralization, making them either unable to fulfill promises or extremely centralized. Ethereum has become the settlement layer for Web3, with user activity happening on layer twos and layer threes that settle back to Ethereum. Zero knowledge rollups use validity proofs that guarantee correctness once submitted to Ethereum, enabling withdrawal periods as short as 10 minutes versus seven days for optimistic rollups. Crucially, zero knowledge rollups do not need to submit transaction data to Ethereum, allowing data to be stored off-chain using data availability layers. This results in fixed proof costs regardless of transaction volume, dramatically reducing fees compared to optimistic rollups where costs increase linearly with data. Public blockchains function like shared servers where popular applications can consume all resources, causing gas fees to spike dramatically. Even during congestion, layer two solutions maintain minimal fees (5-10 cents) compared to Ethereum's $1,000+ fees.

Rollups aggregate multiple transactions into single submissions before sending to the main chain, dramatically improving efficiency. Optimistic Rollups like Arbitrum and Optimism provide the same security as Ethereum but execute transactions off-chain, submitting only results to the main chain. ZK-Rollups use zero-knowledge cryptography to prove transaction validity without revealing data, becoming more efficient as more users join. Unlike Layer 1 networks that slow down with adoption, Rollups improve with user growth. This addresses the fundamental scalability challenge facing Layer 1 blockchains as mass adoption increases.

The London hard fork introduced EIP-1559, fundamentally changing how transaction fees work. The gas limit doubled from 15 million to 30 million per block, allowing more transactions. The new fee structure includes a base fee that burns (destroys) and a priority fee (tip) that goes to miners. This creates deflationary pressure as approximately 1 ether per block is burned. The market immediately reacted to this scarcity, with ether reaching nearly $3000. Miner revenue now correlates directly with network usage—more transactions mean more fees collected.

The speaker explains EIP-1559, which introduces a base fee that dynamically adjusts to move the total gas used in a block toward the current limit of 10 million gas. The base fee is 'burned' (destroyed) rather than given to miners, which makes Ethereum deflationary. Users specify a 'tip' parameter together with the maximum amount they're willing to pay, and miners keep the tip as incentive for inclusion. The speaker notes that this mechanism makes fee estimation more predictable for wallet developers and removes the incentive for miners to manipulate fees. The speaker also mentions that this addresses the problem of high transaction fees that can cost users a significant percentage of their investment.

EIP-1559 introduces a fee burn mechanism where a portion of transaction fees is destroyed rather than going to miners; the video explains that on Etherscan, each block displays both the block reward and the fee burn separately, with the percentage indicating how much of the transaction fees were burned versus remaining as miner tips, and notes that with the gas limit doubled to 30 million, larger blocks show higher burn percentages (around 95% in some cases), resulting in approximately a 20% revenue impact on miners during the initial rollout.

EIP-1559 is an Ethereum Improvement Proposal that introduced a new fee structure for transactions. Under this system, transactions include a base fee that is burned (destroyed) and a priority fee (tip) that goes to miners. This mechanism was designed to make transaction fees more predictable and to transition Ethereum toward a more energy-efficient proof-of-stake consensus mechanism.

EIP1559 introduced a mechanism to burn or permanently destroy the base fee from every transaction. When institutions use the network more, more transactions occur, meaning more ETH gets burned daily. At the same time, new ETH is created to reward stakers. The magic happens when the amount of ETH being burned from transaction fees exceeds the new ETH being paid out. When this occurs, Ethereum becomes deflationary and total supply actually decreases.

Ethereum's scaling roadmap progresses through Proto Danksharding (200 TPS), Full Danksharding (50,000 TPS), and EigenDA (500,000+ TPS, targeting 50 million TPS). In the rollup-centric vision, L2s inherit security but require proof verification (ZK rollups) or fraud proof adjudication (optimistic rollups), plus posting data to a DA service. DA systems function as high-availability bulletin boards, not long-term storage, with data expiring after ~2 weeks. Data blobs attach via transactions with fingerprints, using EIP-1559-style fee mechanics (131,072 gas per blob). Blobs store for 4,096 epochs (18.2 days), yielding 51 GB storage and 32 KB/second throughput. Before 4844, L2s competed with other users for block space; 4844 introduces purpose-built blobs eliminating this competition.

EIP-4844 (Proto-Danksharding) is an Ethereum Improvement Proposal that introduces blob transactions, a new type of transaction that stores large amounts of data in beacon nodes for approximately two weeks before being pruned. This innovation enables Layer 2 solutions like rollups to reduce their operational costs by an order of magnitude by storing transaction data in cheaper blob space rather than expensive call data, thereby dramatically improving Ethereum's scalability while laying the groundwork for future full Danksharding implementation.

Proto-dank sharding is being proposed as a key first step toward full dank sharding implementation. It comes with Ethereum Improvement Proposal 4844, authored by Ethereum researcher Dank Rad Feist and co-founder Vitalik Buterin. The proposal involves adding a new transaction type called blob-carrying transactions to Ethereum. Currently, Ethereum blocks can carry 50-100 kilobytes of data, but under proto-dank sharding, they'll be able to carry closer to one megabyte—a 10 times increase in data availability. While this is less than the 16 megabytes projected under full dank sharding, it should significantly decrease the cost of using roll-ups and pave the way for cheaper, quicker layer 2 networks.

Proto-Danksharding (EIP4844) achieves significant scalability improvements without requiring full sharding infrastructure. After the merge separates Ethereum into consensus and execution layers, data is stored temporarily (for months, not indefinitely) before pruning. This bounded amount of data can be distributed among different validator nodes, providing enough time for layer 2 protocols to secure their networks while avoiding infinite data growth. Current Ethereum blocks are 50-100 gigabytes with call data available forever; Proto-Danksharding increases blob data to approximately 1 megabyte per block (order of magnitude improvement). Full Danksharding can achieve another order of magnitude by distributing data across 64 shard nodes. These two markets (computation gas and data storage) can be separated using multi-dimensional EIP-1559 mechanisms. Trusted setups for cryptographic schemes like KZG commitments require generating elliptic curve points with specific mathematical relationships, where nobody should know the actual secret relation. Multi-party computation distributes trust among many participants such that even if n-1 participants colluded, they couldn't reconstruct the secret if one participant properly destroyed their data.

Proto-danksharding, implemented in EIP-4844, is the mechanism for dividing the Ethereum blockchain into smaller shards. Blobs are collections of data stored off-chain and attached to blocks, serving as the first step toward fragmentation. Unlike current persistent memory (call data) which must be maintained by every node forever and is slow and expensive, blobs are stored temporarily and automatically deleted after two weeks, significantly reducing storage requirements. The update introduces blob transactions and a multi-dimensional fee market to properly price blobs.