The Truth Behind the L2 Scaling Debate: Why Rollup‑First Is the Best Path Today
3-Point Summary
- Rollup-first is Ethereum’s most practical and powerful scalability path today.
- L1 focuses on security and data availability, while L2 Rollups handle execution and scalability.
- Blob Space and commitments allow L2 data growth without increasing L1’s long-term burden.
Ethereum’s scalability doesn’t come from adding more L2s — it comes from the Rollup-first model that cleanly separates L1 security from L2 execution.
20-Second Shorts Video
“Rollup-first is the best path for Ethereum’s scalability today”
As Vitalik has emphasized many times, Ethereum’s current scalability strategy is firmly based on a Rollup-first approach.
In this model, L1 focuses on security and data availability, while L2 (Rollups) handle execution and scalability — a division of responsibilities that represents the most practical and powerful path forward at this stage.
This article is best understood when read together with two previous pieces.
First, the broader direction of digital financial infrastructure is discussed in
“The Future of Digital Financial Infrastructure: Why L1 Fragmentation Fails and L2 Becomes the Best Business Model”
,
and the security differences between L1 and L2 are analyzed in
“Is L2 Really Safe? The Critical Security Differences Between L1 and L2”
.
As the Ethereum ecosystem grows, criticisms such as “There are too many L2s,” “L2s are basically separate chains,” or “Posting data to L1 increases its burden” continue to surface.
However, these concerns stem from misunderstandings about how Ethereum L1 is structured and how Rollups are designed.
This article breaks down the actual structure of L1, how Blob space works, and how Rollups achieve scalability — and then examines the conceptual direction Ethereum aims for in the long term.
1) Ethereum L1 consists of five components
Understanding this structure makes it clear why L1 does not become heavier even when many L2s post data to it.
① Smart contract code (Code)
The program logic that defines contract behavior. It is stored permanently on-chain.
② Smart contract state (State)
Balances, variables, and storage — the heaviest data that L1 must maintain permanently.
③ Blocks
The ledger of transactions. Block headers include metadata and blob commitments.
④ CALLDATA
A read-only space containing transaction input data. It does not increase state.
⑤ Blob Space (EIP‑4844)
A temporary storage area for large batches of rollup data. L2s submit transaction bundles as Blobs, and L1 does not store the full Blob content in the block. Instead, the Blob body is stored in Blob Space, while the block header stores only a commitment proving the Blob has not been tampered with.
This design allows L1 to guarantee “data existence and integrity” while minimizing storage burden. Blob bodies are deleted after a certain period, but commitments remain permanently in block headers, proving the data existed exactly as posted.
In short:
Blob = data body,
Commitment = fingerprint of the data,
Block header = receipt containing that fingerprint.
2) Why L1 does not become heavier even when many L2s post data
① L1 does not execute or store rollup data as state
Rollup data does not increase L1 state. L1 only records that the data exists.
② Blob Space + commitment structure keeps L1 overhead extremely low
The data body is stored in Blob Space (temporary storage),
L1 stores only a small commitment,
State does not increase at all.
③ Ethereum’s roadmap is designed around increasing L2 data
The Danksharding roadmap explicitly aims to expand Blob capacity.
Thus, “More L2 data = L1 overload” is structurally incorrect.
3) Why rollup state can be reconstructed even after Blobs are deleted
EIP‑4844 introduced Blob Space, where rollup data is stored temporarily. Blob bodies are deleted after a certain period. Can rollup state still be reconstructed afterward? The answer is yes.
① L1 permanently stores the Blob’s authenticity, not its body
Blob bodies are deleted, but their commitments remain permanently in block headers. These commitments act as fingerprints proving the data was correct at the time of posting.
② Blob data is stored across the network
While L1 does not store Blob bodies permanently, L2 sequencers, validators, archive nodes, DA layers (e.g., EigenDA), and even regular users can store copies. Data retention is distributed across the network, not centralized on L1.
③ Only two things are needed for reconstruction: data body + commitment
To reconstruct rollup state, the Blob body is needed. To verify its authenticity, the L1 commitment is sufficient. Anyone can retrieve the Blob data from the network and compare it to the L1 commitment to confirm it is the original.
In short:
Blob Space = temporary upload area,
Commitment = permanent fingerprint proving correctness,
Network = actual data holder.
4) Native Rollup is still a conceptual model
Native Rollup is not a fully implemented technology today. It is a conceptual model describing the direction Ethereum aims for in the long term. It builds on the security and data structure of current Rollups, but the following points represent the intended direction rather than present reality.
① It aims to inherit L1 security directly
All rollup data is posted to L1, and state transitions are verified under L1 consensus.
ZK Rollups rely on L1 verifiers, and Optimistic Rollups rely on L1 fraud proofs.
Trust is placed in L1 code, not operators. The model assumes rollup security equals L1 security.
② It treats rollups as L1 extension layers, not separate chains
Native Rollup assumes no separate consensus or validator set.
Rollup state can be fully reconstructed from L1 data, and finality is determined by L1.
Execution is separated, but security remains unified.
③ It aims to solve L2 fragmentation at the root
All L2s share the same L1 data availability layer.
Data, proofs, and finality are all anchored to L1, enabling natural interoperability.
With real-time proving, multiple L2s could behave like a single synchronous system.
The long-term vision is a unified “Ethereum experience” regardless of which L2 a user interacts with.
In summary, Native Rollup is a future-oriented model that encapsulates Ethereum’s long-term philosophy for security, consensus, and data architecture — not a fully realized system today.
Conclusion
Ethereum’s scalability strategy is not about creating many L2s, but about a clear division of responsibilities:
L1 focuses on security and data availability, while L2 focuses on execution and scalability.
Blob Space minimizes L1’s storage burden while guaranteeing data integrity,
and Rollups inherit L1 security, allowing state to be reconstructed at any time.
This structure reflects Ethereum’s core philosophy: “L1 should do the minimum; L2 should scale the system.”
It enables the network to expand throughput without making L1 heavier.
Ultimately, Rollup-first is the best scalability path for Ethereum today,
and the clear separation between L1 and L2 is what makes the ecosystem both stable and sustainable.
Younchan Jung
Researcher exploring structural shifts in AI, blockchain, and the on‑chain economy.
If you would like to read this article in Korean, please click the button below.