TL;DR: The Fusaka upgrade, launching late 2025, merges the Fulu and Osaka hard forks to scale Ethereum’s data capacity. By implementing PeerDAS and Verkle Trees, Fusaka stabilizes Layer 2 (L2) fees at $0.005 and enables stateless nodes, removing the 2TB storage requirement for validators.
Who This Is For
Fusaka targets Ethereum developers requiring high-throughput data availability, institutional operators seeking lower hardware overhead, and retail users demanding sub-cent transaction costs without the security risks of seed phrases.
Our Verdict
Fusaka is the most significant structural overhaul since The Merge. It effectively ends the "high gas" era for L2s and eliminates the hardware centralization risks inherent in scaling. This upgrade transforms Ethereum from a high-cost settlement layer into a high-performance, invisible utility backend.
The Fusaka upgrade—comprising the Fulu (Consensus) and Osaka (Execution) hard forks—re-engineers Ethereum’s data and state management. It delivers the final technical requirements for The Surge (scalability) and The Verge (statelessness), shifting the network toward a future defined by biometric security and invisible infrastructure.
1. The Economics of Sub-Cent Fees
Fusaka pushes Layer 2 (L2) transaction fees consistently below $0.01. Network data from December 2025 shows gas fees on Arbitrum, Optimism, and Base have stabilized at approximately $0.005. This efficiency allows Ethereum to match the throughput of "Alt-L1" networks like Solana while retaining superior decentralization.
EIP-7918 drives this economic shift by linking blob fees directly to L1 execution costs. This mechanism ensures L2 activity fuels the ETH burn. Post-activation data indicates that blob fees account for up to 98% of daily ETH burned during peak activity, reinforcing the "ultrasound money" thesis despite lower user costs.
2. PeerDAS: Scaling Data Capacity
PeerDAS (Peer Data Availability Sampling) removes the primary bottleneck for scaling rollups: the requirement for every validator to download 100% of "blob" data. Instead, validators verify data integrity by sampling random portions from their peers.
- Bandwidth: PeerDAS cuts validator bandwidth requirements by 85%.
- Throughput: This optimization expands the blob target from 6 to over 32, supporting network capacity exceeding 100,000 transactions per second (TPS).
PeerDAS facilitates higher data throughput without pricing out home stakers. By distributing the data burden, we prevent validator centralization into specialized data centers. — Kenji, Security Researcher
3. Verkle Trees: Eliminating the 2TB Storage Barrier
Fusaka replaces Merkle Patricia Tries with Verkle Trees. Using polynomial commitments, Verkle Trees shrink "witness" sizes—the cryptographic proofs needed to verify blocks—from 150 KB to 2 KB. This transition enables Statelessness.
Stateless clients allow nodes to verify the blockchain without storing the 2TB+ state history. This democratization allows mobile devices and low-power hardware to function as full nodes, significantly hardening the network against centralized points of failure.
4. EVM Evolution: EOF and Gas Expansion
The Osaka upgrade introduces the EVM Object Format (EOF). EOF provides a versioned container for smart contract code, separating code from data. This structure simplifies code optimization and mitigates several common smart contract vulnerabilities.
Simultaneously, Fusaka increases the block gas limit to 60 million. This expansion widens the execution pipeline to accommodate increasing L2 proofs and institutional Real World Asset (RWA) transactions, with a long-term target of 150 million.
5. Consumer Impact: Biometric Onboarding
EIP-7951 introduces WebAuthn (secp256r1) support. This allows users to sign transactions via native hardware security modules, such as Face ID or fingerprint sensors. This eliminates the need for seed phrases and enables a "key-less" user experience where applications can seamlessly sponsor user fees.
Key Takeaways
- Sub-Cent L2 Fees: Transaction costs stabilize at $0.005, making microtransactions viable.
- PeerDAS: Increases data capacity 8x while reducing bandwidth overhead by 85%.
- Statelessness: Verkle Trees reduce witness sizes by 98%, removing the need for 2TB local storage.
- WebAuthn: Biometric signing replaces complex seed phrases for mainstream onboarding.
- Economic Alignment: EIP-7918 ensures L2 growth continues to drive ETH deflation via blob fee burning.
Would you like me to generate a comparison table of Ethereum node storage requirements before and after Verkle Tree implementation?
