The Core Concept
As of late May 2023, Ethereum is in the midst of one of the most significant architectural transformations in its history. The network’s transition from a monolithic blockchain — where every node processes every transaction — to a rollup-centric model represents a fundamental rethinking of how blockchain scalability can be achieved. With Bitcoin trading at $28,086 and Ethereum at $1,911, the crypto market is showing renewed strength, but the real story for Ethereum lies not in price action but in the technical infrastructure being built beneath the surface.
The rollup-centric roadmap, formally adopted by Ethereum core developers in late 2022 following the Merge, prioritizes scaling through layer-2 protocols called rollups rather than trying to increase the base layer’s throughput directly. Rollups bundle hundreds or thousands of transactions off-chain, compress them, and post a single cryptographic proof back to the Ethereum mainnet. This approach inherits Ethereum’s security guarantees while dramatically increasing throughput — theoretically enabling the network to process tens of thousands of transactions per second without sacrificing decentralization.
The next critical milestone in this roadmap is EIP-4844, also known as proto-danksharding, which introduces a new transaction type called “blob-carrying transactions.” These blobs provide a dedicated, cost-efficient data storage mechanism for rollups, reducing their operating costs by an estimated 10x to 100x compared to current calldata-based approaches.
How It Works Under the Hood
Proto-danksharding’s technical innovation lies in how it handles data availability. Currently, rollups like Arbitrum, Optimism, Polygon zkEVM, and zkSync Era post their transaction data to Ethereum as calldata — essentially permanent, on-chain storage that is processed by every Ethereum node. This is expensive. Every byte of calldata costs 16 gas, and with mainnet gas prices frequently spiking during periods of high demand, rollup operators pass these costs on to users in the form of elevated fees.
EIP-4844 introduces blob transactions that store data in a separate, temporary space. Blobs are not accessible by the Ethereum Virtual Machine — they exist purely for data availability purposes. This means rollups can post their transaction data to blobs at a fraction of the cost, since the data does not need to be permanently stored or processed by smart contracts. The blobs are pruned from the blockchain after approximately 18 days, keeping node storage requirements manageable.
The cryptographic backbone of this system is the KZG polynomial commitment scheme. Unlike Merkle trees, which require storing intermediate hashes to prove data inclusion, KZG commitments allow verifiers to confirm that a specific piece of data exists within a blob using a constant-size proof. This is computationally efficient and enables constant-time verification regardless of the blob’s size. The Ethereum community conducted a trusted setup ceremony in early 2023 to generate the cryptographic parameters needed for KZG commitments, with thousands of participants contributing entropy to ensure the system’s security.
The upgrade also lays the groundwork for full danksharding, Ethereum’s long-term scaling solution. Full danksharding will expand the blob space dramatically, potentially supporting hundreds of parallel blob-carrying transactions per block. Proto-danksharding is deliberately designed as a stepping stone — it implements the same transaction format and fee market structure that full danksharding will use, but with a smaller initial blob capacity.
Real-World Applications
The practical impact of proto-danksharding extends across the entire Ethereum ecosystem. Layer-2 protocols like Arbitrum and Optimism, which already process millions of transactions per month using optimistic rollup technology, stand to benefit immediately from reduced data posting costs. Zero-knowledge rollups — including zkSync Era, StarkNet, Polygon zkEVM, and Scroll — are equally positioned to leverage blob transactions for their proof data.
For decentralized finance applications, the cost reduction could be transformative. Current DeFi protocols on Ethereum layer-2 networks charge users between $0.01 and $0.50 per transaction depending on network congestion. With blob transactions, these fees could drop to fractions of a cent, making micropayments, automated trading strategies, and high-frequency DeFi interactions economically viable for the first time. The implications extend to gaming, social media protocols, and supply chain tracking — applications that have been constrained by the cost of on-chain interactions.
Cross-chain bridges and interoperability protocols also benefit. Many bridges rely on posting verification data to Ethereum mainnet, and the current calldata costs make frequent state updates expensive. Blob transactions could enable real-time cross-chain communication with significantly lower overhead, improving the user experience for multi-chain DeFi strategies and asset transfers.
Scalability and Limitations
Despite its promise, proto-danksharding is not a silver bullet. The initial implementation targets a modest increase in blob capacity — approximately one to two blobs per block, each containing around 128 kilobytes of data. While this represents a meaningful cost reduction for rollups, it falls far short of the throughput needed to support billions of daily transactions. Full danksharding, which would scale blob capacity to hundreds per block, remains years away and depends on further advances in peer-to-peer networking and data availability sampling.
The temporary nature of blob storage also introduces new considerations for rollup design. Current rollup architectures assume that transaction data is permanently available on Ethereum. With blobs being pruned after approximately 18 days, rollups need to ensure that their data is reconstructed and stored independently before it disappears from the base layer. This requires additional infrastructure — such as decentralized storage networks or independent archival nodes — adding complexity to systems that already struggle with centralization concerns around their sequencer components.
There are also open questions about the fee market dynamics of blob transactions. EIP-4844 introduces a separate gas market for blobs, with its own base fee and priority fee structure. If demand for blob space exceeds the available capacity, blob fees could spike independently of regular Ethereum gas prices — potentially creating a situation where rollup costs are driven by blob congestion rather than general network activity. The economic modeling around this dual fee market is still largely theoretical.
The Future Horizon
Looking ahead, proto-danksharding represents the beginning of Ethereum’s most ambitious technical chapter. The upgrade’s modular design means that subsequent improvements — larger blob sizes, increased blob counts, and improved data availability sampling — can be rolled out incrementally through hard forks without requiring changes to the core transaction format. This forward-compatibility is a deliberate design choice that protects the ecosystem from disruptive protocol changes.
The competitive landscape is also driving urgency. Alternative layer-1 blockchains like Solana, which processes thousands of transactions per second natively, and app-specific chains built on the Cosmos SDK continue to attract users with their low fees and high throughput. Ethereum’s bet on the rollup-centric approach must deliver tangible cost savings and performance improvements to maintain its position as the dominant smart contract platform.
With the broader crypto market showing signs of recovery — Bitcoin at $28,086 representing a 5% weekly gain, and Ethereum at $1,911 reflecting similar momentum — the timing for proto-danksharding’s implementation is favorable. A more scalable Ethereum base layer could unlock the next wave of on-chain innovation, bringing decentralized applications closer to mainstream adoption. The technical foundation is being laid; the question now is how quickly the ecosystem can build on it.
Disclaimer: This article is for informational purposes only and does not constitute financial advice. Cryptocurrency markets are highly volatile, and readers should conduct their own research before making any investment decisions. Past performance is not indicative of future results.
proto-danksharding is the most important ETH upgrade since the merge. blob transactions will change everything for L2 fees
tens of thousands of TPS without sacrificing decentralization. thats the promise anyway. still waiting to see it in practice
andrei is right to be skeptical. every ETH scaling promise has been 18 months away for the last 5 years
18 months away stopped being a joke when EIP-4844 actually shipped. L2 fees went from $5 to $0.02. the scaling roadmap is working, just slower than everyone wanted
the 5 dollar to 2 cent fee drop after 4844 was the moment L2s became actually usable. people forget how bad it was before blobs
rollups are doing 2000+ TPS right now on base and arbitrum. the tens of thousands number needs sharding which is still years out. but the current throughput is real and usable
base doing 2000 TPS right now is proof the roadmap works. sharding would be nice but L2 throughput is already enough for most use cases