Ethereum’s upgrade roadmap has executed on schedule better than almost any large-scale open-source software project in history. The Merge (September 2022) cut energy consumption by 99.95%. Shapella (April 2023) unlocked staked ETH withdrawals. Dencun (March 2024) reduced L2 fees by over 90%. Each upgrade was controversial before launch and quickly accepted once live. Understanding what’s already shipped, and what’s next, matters for anyone building on or investing in Ethereum.
What was the Ethereum Merge and why did it matter?
The Merge, executed on September 15, 2022, transitioned Ethereum from proof-of-work (PoW) to proof-of-stake (PoS). It was one of the most technically complex live system migrations in software history, essentially replacing the engine of a running aircraft. The network continued producing blocks without interruption.
What actually changed:
- Energy consumption: Dropped ~99.95% overnight. Ethereum’s annualized energy use went from roughly equivalent to a mid-sized country to comparable to a small city.
- Issuance rate: ETH issuance dropped ~88% post-Merge, combined with EIP-1559’s fee burning this made ETH periodically deflationary during high-activity periods.
- Mining eliminated: GPU miners were replaced by validators who stake 32 ETH as collateral. This removed the constant sell pressure from miners covering electricity costs.
- Security model: Attacking the network now requires acquiring 33%+ of staked ETH (currently over 32 million ETH staked), making attacks economically prohibitive.
What the Merge did not do
A common misconception: the Merge did not make Ethereum faster or reduce gas fees. Throughput remained roughly the same (~15-30 TPS on mainnet). The scalability improvements came later, through L2 rollups and the Dencun upgrade. This is why gas fees didn’t drop in September 2022, that wasn’t what the Merge was designed to fix.
What was the Beacon Chain and why did Ethereum need it?
The Beacon Chain launched December 1, 2020, nearly two years before the Merge. It ran the PoS consensus layer in parallel with the existing PoW chain, allowing the system to be tested under real conditions with real validators and real staked ETH before becoming the canonical chain.
Validators had to deposit 32 ETH (one-way, until Shapella) to participate. Over 500,000 validators participated by the time of the Merge, collectively staking over 16 million ETH. This parallel testing period gave the Ethereum team high confidence in the PoS system’s stability before committing the main chain to it.
What did the Dencun upgrade do for Ethereum in 2024?
Dencun deployed on March 13, 2024 and was the most impactful upgrade for everyday Ethereum users since the Merge. Its key feature: Proto-Danksharding (EIP-4844), which introduced “blob” transactions, a new data storage mechanism specifically designed for L2 rollups.
The practical impact was immediate. L2 transaction fees dropped 90%+ within days:
- Arbitrum One average fees dropped from ~$0.50 to ~$0.03
- Optimism fees dropped similarly
- Base (Coinbase’s L2) became practically free for many transaction types
How it works: instead of posting L2 transaction data as calldata on the Ethereum mainchain (expensive, permanent), rollups now post it as blobs, temporary storage that’s cheaper because it’s automatically pruned after ~18 days. The data is still available long enough to verify rollup integrity, just not stored forever.
What is the Pectra upgrade and what does it change?
Pectra (combining Prague and Electra upgrade names) is the next major Ethereum upgrade following Dencun. It packages multiple EIPs with meaningful user-facing improvements:
- EIP-7702: Account Abstraction, allows regular wallet addresses (EOAs) to execute code, enabling smart wallet features like batched transactions, sponsored gas, and social recovery without migrating to a smart contract wallet
- EIP-7251: Raises the maximum effective balance for validators from 32 ETH to 2,048 ETH, reducing the number of validators needed and improving network efficiency
- EIP-7549: Improves attestation efficiency, reducing bandwidth requirements for validators
- Blob count increase: Further expanding blob capacity from Dencun’s baseline, continuing to reduce L2 costs
Account abstraction (EIP-7702) is the most significant for end users: it enables transaction batching (approve + swap in one transaction), removes the need to hold ETH for gas (dapps can sponsor fees), and allows wallet recovery mechanisms beyond seed phrases. This is the UX improvement the industry has been waiting for.
How does the Ethereum Layer 2 ecosystem work?
Layer 2 solutions process transactions off the Ethereum mainchain but post proof of those transactions back to Ethereum, inheriting its security. Two main types:
- Optimistic rollups (Arbitrum One, Optimism, Base): Assume transactions are valid unless challenged. Have a ~7-day withdrawal delay to mainnet. Simple to implement, widely adopted.
- ZK rollups (zkSync Era, Starknet, Polygon zkEVM, Scroll): Use zero-knowledge proofs to cryptographically verify transaction batches. Faster finality, but more complex. The long-term scaling path Ethereum’s roadmap targets.
Post-Dencun, both types now use blobs for data availability, dramatically reducing costs. The Ethereum roadmap (“The Surge”) targets 100,000+ TPS at the L2 layer with full Danksharding, a multi-year project that would make blob space abundant and L2 costs approaching zero.
How does Ethereum staking work in 2025-2026?
Post-Shapella (April 2023), staked ETH can be withdrawn, removing the main liquidity concern. Options for staking:
- Solo staking: 32 ETH minimum, run your own validator node. Maximum rewards and decentralization, but requires technical management.
- Liquid staking (Lido, Rocket Pool, Frax): Stake any amount, receive liquid staking tokens (stETH, rETH) that can be used in DeFi. Lido holds ~30% of all staked ETH, raising centralization concerns the community actively monitors.
- Exchange staking (Coinbase, Kraken): Simplest setup, lowest yield after platform fees, least decentralized.
Current staking APR: approximately 3.5–5% annually for solo/liquid staking, varying with network activity. Pectra’s increase to 2,048 ETH max balance will reduce the total validator count without reducing total staked ETH, potentially improving validator efficiency and slightly increasing APR for those who can run larger stakes.
Frequently Asked Questions
Did the Ethereum Merge make transactions cheaper?
No, the Merge didn’t reduce gas fees. It addressed energy consumption and issuance, not throughput. The fee reduction came from the Dencun upgrade (March 2024) which enabled blob transactions, cutting L2 fees by 90%+. Today, most Ethereum users are encouraged to transact on Layer 2 networks like Arbitrum, Optimism, or Base rather than mainnet to avoid high gas costs.
Is Ethereum still being upgraded after Dencun?
Yes. The Pectra upgrade follows Dencun and focuses on account abstraction (EIP-7702), validator efficiency (EIP-7251), and expanded blob capacity. After Pectra, the Ethereum roadmap continues with “The Surge” (full Danksharding), “The Scourge” (MEV resistance), “The Verge” (Verkle trees for stateless clients), and “The Purge” (historical data reduction). Ethereum’s development is ongoing, with a multi-year roadmap.
How many ETH do you need to stake on Ethereum?
Solo validator operation requires 32 ETH. Liquid staking protocols like Lido accept any amount, you deposit ETH and receive stETH (a liquid token representing your staked position) in return. After the Pectra upgrade, the maximum effective balance per validator increases to 2,048 ETH, but the 32 ETH minimum for solo staking remains the same.
