The Ethereum network's Dencun upgrade activated on March 13, 2024 at epoch 269568, introducing EIP-4844 (proto-danksharding) — the most consequential scaling improvement to Ethereum since the Merge. Over a year later, the data is in, and it tells a more nuanced story than the initial headlines suggested.
What EIP-4844 Actually Does
Before Dencun, Layer 2 networks (Arbitrum, Optimism, Base, zkSync, Starknet, and others) posted their transaction data to Ethereum as calldata — a permanent, expensive form of data storage billed at Ethereum's standard gas rates. This was the single largest cost driver for L2 operators, and it was passed directly to users as transaction fees.
EIP-4844 introduced a new transaction type carrying blobs — large data packets (~128 KB each) attached to blocks but stored only temporarily (roughly 18 days). Blobs have their own fee market, independent of Ethereum's main execution gas market, and they are not accessible to the EVM — they cannot be read by smart contracts.
Why Temporary Storage Is Sufficient
Layer 2 networks use Ethereum for data availability — to prove that their off-chain transactions actually occurred. Once a fraud proof window closes (typically 7 days for optimistic rollups) or validity proofs are submitted (ZK rollups), the underlying data no longer needs to be stored on-chain. Blobs are sized precisely for this use case.
The Fee Impact: Real Numbers
The reduction in L2 fees was immediate and significant. Based on data from L2Fees.info and Dune Analytics dashboards tracking the month before and month after the Dencun activation:
Approximate average swap transaction fees, 30-day period before vs. 30-day period after Dencun activation. Source: L2Fees.info, Dune Analytics.
For context, a $0.02–0.04 transaction fee positions Ethereum L2s competitively against Solana (~$0.001–0.003) and significantly cheaper than Ethereum mainnet (typically $1–15 for a swap, depending on congestion).
What Changed for DeFi Users
The fee reduction unlocked use cases that were economically unviable on Ethereum mainnet and even borderline on pre-Dencun L2s:
Micro-transactions and gaming. Applications requiring dozens of on-chain interactions (blockchain games, loyalty programs, prediction markets with frequent small positions) became feasible for the first time on Ethereum-secured infrastructure.
Dollar-cost averaging at scale. Recurring DeFi operations — weekly DCA swaps, auto-compounding yield strategies, rebalancing — that cost $2–5 per execution pre-Dencun now cost less than $0.10, changing the math on small-balance participation significantly.
Cross-L2 arbitrage efficiency. Lower fees reduced the spread required to make cross-L2 arbitrage profitable, which theoretically improves price consistency across the L2 ecosystem.
What EIP-4844 Did Not Solve
Incomplete Picture
The dramatic fee reduction figures above reflect optimal conditions — periods of low blob demand. The blob fee market introduces a new form of congestion that materialized during high-activity events in 2024. Understanding this is essential context.
Blob congestion. Ethereum blocks currently support a target of 3 blobs and a maximum of 6 per block. During peak activity events (major token launches, network stress tests, high L2 volume), blob space saturated and blob fees spiked by 100–500x within hours. During these episodes, L2 fees temporarily returned to pre-Dencun levels or higher.
Finality latency. EIP-4844 does not change Ethereum's finality timeline. Optimistic rollup transactions still require a 7-day challenge window for full settlement on L1. For many institutional use cases requiring fast finality, ZK rollups remain the relevant architecture.
Full danksharding is not yet live. EIP-4844 is explicitly "proto-danksharding" — a stepping stone to the full danksharding roadmap, which would increase blob capacity by an order of magnitude (64 blobs per block vs. the current 6 maximum). Full danksharding requires PeerDAS (Peer Data Availability Sampling) and other infrastructure that remains in research and development as of mid-2025.
What Comes Next
The Ethereum development roadmap following Dencun includes Pectra (scheduled for late 2025), which includes EIP-7691 to increase the blob count targets, and early-stage work on PeerDAS that moves Ethereum toward the full danksharding design.
For L2 users, the practical implication is that the fee environment should continue improving incrementally — but the dramatic single-event reduction seen at Dencun is unlikely to repeat. Future improvements will be more gradual, tied to the pace of blob capacity expansion and ZK proof system maturation.
CoinLens Assessment
EIP-4844 delivered on its core promise: it meaningfully reduced the cost of using Ethereum-secured Layer 2 networks for everyday transactions. The 90%+ fee reductions are real and represent a genuine improvement in Ethereum's scalability profile.
However, the blob fee market introduces a new congestion variable that users and developers need to model, and the full danksharding vision that would make blob congestion a non-issue is still multiple upgrade cycles away. Ethereum's scaling roadmap is on track — but it remains a multi-year project.
