Ethereum has always aimed to be credibly neutral while letting innovation flourish on higher layers. Early discussions sketched a “rollup-centric roadmap” in which the base layer would simplify and ossify so that most activity could migrate to L2s. Recent developments, however, have recognized that being a minimal consensus and data availability layer is not enough: L1 should be able to handle traffic and activity since that is ultimately what L2s are plugging into. This means faster blocks, cheaper data, stronger proofs, and better interoperability.

The upcoming Beam Chain consensus redesign targets faster finality and lighter validator requirements, reinforcing Ethereum’s neutrality while boosting raw throughput. In parallel, proposals are being considered to migrate activity from the aging (and “complexifying”) EVM to a RISC-V native VM, promising massive gains in prover efficiency while preserving interoperability with legacy contracts.

These upgrades will reshape the L2 landscape. By 2030, I see Ethereum’s general rollup-centric roadmap consolidating toward two poles on a spectrum:

• Aligned rollups: Prioritize deep Ethereum integration (e.g., shared sequencing, native verification) and fully leverage liquidity on L1 with minimal trust assumptions. The relationship is mutually beneficial with aligned rollups leveraging composability and security directly from L1.

• Performance rollups: Prioritize throughput and real-time UX, sometimes via alternative DA layers or permissioned actors (e.g., centralized sequencer, small security councils/multisigs), yet still settle to Ethereum for credibility (or marketing).

In designing these rollups, every team will need to navigate three tradeoffs:

1. Liquidity access: How to access and use liquidity on Ethereum and, potentially, other rollups? How important is synchronous or atomic composability?

2. Security source: To what extent should the liquidity from Ethereum that is moved onto the rollup inherit security directly from Ethereum or rely on the rollup provider?

3. Execution expressiveness: How important is EVM compatibility? Will EVM-compatibility continue to be important in five years time given the rise of alternatives like SVM and popular Rust smart contracts?

Answering these questions will determine whether a rollup gravitates more toward alignment vs. performance (and also are relevant to being community-operated vs corporate-controlled). In the rest of this report, I’ll explore how Ethereum’s forthcoming protocol upgrades power both rollup archetypes, and why those in the middle risk obsolescence.

Rollups in the top-left of the chart above emphasize performance: they might employ centralized sequencers, alternative DA networks, or application-specific optimizations to achieve far higher throughput than the typical L2 (e.g., MegaETH). Some performance rollups will fall further right on alignment (e.g., by leveraging fast based preconfs like Puffer UniFi and Rise; targeting the holy grail in the top-right corner), but hard finality will still be dependent on L1 specs. In contrast, rollups in the bottom-right maximize Ethereum alignment: implementing ETH deeply across fees, exchange, and DeFi; enshrining their transaction ordering and/or proof verification on L1; and favoring composability over raw speed (e.g., Taiko has been trending in this direction, but they are also exploring permissioned preconfs for better UX now). By 2030, I expect many “middle-of-the-road” L2s to either pivot toward one of these archetypes or risk being outcompeted. Users and developers will gravitate to either high-security, Ethereum-aligned environments for high-stakes and composable DeFi, or high-scalability, app-tailored networks for mass-user applications. Ethereum’s roadmap toward 2030 lays the foundation to support both tracks.

Network effects push markets toward fewer, bigger hubs. In markets like crypto where network effects play a major role, the market will likely end up dominated by a few winners (like we see with CEXs as well). Because network effects coalesce around a chain’s strongest attributes, ecosystems will tend to consolidate around a few performance-maximal and security-maximal platforms. A rollup that only half-commits to Ethereum alignment or performance might end up with neither the security of the former nor the usability of the latter.

As rollup technology matures, economic activity will stratify based on the tradeoff between security needed versus cost to acquire that security. Use-cases that cannot tolerate settlement or governance risk – institutional DeFi, large onchain treasuries, high-value collateral markets, etc. – will likely concentrate on chains that inherit Ethereum’s full security guarantees and neutrality (or Ethereum L1 itself). On the other end, the most mass-adoption activity (memes, trading, social, gaming, retail payments, etc.) will concentrate on chains with the best UX at the lowest cost, which may require custom throughput boosts or centralized sequencing. Thus, there is diminishing appeal for a generic chain that is somewhat fast but not fastest, or somewhat secure but not the most secure – especially if by 2030, cross-chain interoperability makes it trivially easy to move assets between these two worlds.

Ethereum’s current execution environment (the EVM with its 256-bit architecture and legacy design) may be replaced or augmented with a more modern, efficient VM by 2030. Vitalik has proposed transitioning Ethereum’s VM to a RISC-V based architecture. RISC-V is a streamlined, modular instruction set that promises major efficiency gains (up to 50-100x) in transaction execution and proof generation. It uses 32/64-bit instructions which are directly compatible with modern CPUs and more efficient for zk proving. To be less disruptive and avoid the risk of stalling progress (e.g., like previously when the Ethereum community considered replacing the EVM with eWasm) the plan is to adopt a dual-VM model, running both the EVM for backward compatibility and a new RISC-V VM for new contracts (similar to how Arbitrum Stylus does this with WASM + EVM contracts already). The goal is to drastically simplify and speed up the execution layer, benefiting both L1 scalability and rollup support.

Why do this?

The EVM wasn’t designed with zk proofs in mind. zk-EVM provers incur heavy overhead when emulating state transitions, computing roots/hash trees, and handling EVM-specific quirks. By contrast, a RISC-V VM replaces this with a simpler, register-based logic that can be directly modeled and proven with far fewer constraints. It is far more zk friendly, eliminating inefficiencies (like gas accounting and state management overhead). This can materially help any rollup that uses zk proofs: proving their state transitions becomes easier, faster, and cheaper. The bottom line is that upgrading the EVM to a RISC-V VM can boost overall proof throughput, making it feasible for L1 to directly verify L2 executions (discussed further below), while also improving the throughput ceiling for performance rollups’ own VMs. 

On top of that, it would also meaningfully expand Ethereum’s dev ecosystem beyond the Solidity/Vyper niche to larger Rust, C/C++, Go, and other mainstream dev communities. 

Ethereum’s proposed move from ad-hoc L2 settlement to a unified, enshrined settlement framework will overhaul rollup settlement as we know it. Today, each rollup deploys its own L1 verification contracts (fraud proofs or validity proofs), which are bespoke and disconnected. By 2030, Ethereum may integrate a native function (the proposed EXECUTE precompile) that acts as a universal L2 execution verifier. EXECUTE allows Ethereum validators to directly re-execute a rollup’s state transition and confirm if it’s correct. In essence, Ethereum would “enshrine” the ability to verify arbitrary rollup blocks within its own protocol.

In effect, this upgrade will enable native rollups that are effectively programmable execution shards (akin to what NEAR does). Native rollups are distinct from L2s, rollups, and based rollups in that Ethereum’s own execution engine verifies their blocks.

EXECUTE significantly simplifies development of EVM-equivalent rollups by removing the need for complex custom infrastructure (e.g., fraud proof games, zk circuits, multisig “security councils”) used for EVM emulation and maintenance. The result is fully trustless L2s with minimal custom codes. Paired with next-gen real-time provers (e.g., Fermah, Succinct), you get real-time settlement natively on L1 (rollup txs achieving finality on L1 as soon as they’re included, without waiting for fraud proof windows or multi-slot proof computations). By making the settlement layer global and shared, Ethereum enhances its credible neutrality (freedom to choose among verifier clients) and composability (synchronous composability is much simpler without worrying about same-slot real-time proving). All native (or native + based rollups) would use the same L1 function to settle, enabling standardized proofs and easier interactions among these rollups (shards).

Ethereum’s Beacon Chain consensus layer is being reimagined from the ground up as Beam Chain (testing by 2027-2029). The aim is to overhaul consensus with advanced cryptography (including quantum resistance) and greater scalability and decentralization. Among various upgrades spanning six dedicated research tracks, I’ve selected a few key features that I see as most relevant to this report:

Regular Beam Chain updates can be followed on the “Beam Call” series on YouTube.

1. Shorter slots, faster finality: One of Beam Chain’s primary goals is faster finality: moving from ~15-min finality today (Gasper’s 2 epochs; 32 + 32 slots at 12-second slot time) to 3-slot finality (3SF) with 4-second slot time (4SST) (~12 seconds) and eventually single-slot finality (SSF; ~4 seconds). 3SF+4SST means that your transaction would be finalized in under 10 seconds after it appears onchain. This would be a big UX improvement for based and native rollups because faster L1 blocks means faster based and native rollup blocks. It would take ~4 seconds to include your tx in a block (or more under high network load), which is a 3x block time improvement for based and native rollups (though that’s still relatively slow compared to performance rollups, alt-L1s, and credit cards, so preconfirmations will likely still be important). Faster L1 finality also helps secure and accelerate settlement – rollups would be able to get their state commits finalized on the base layer in seconds, making withdrawals rapid and closing the window for reorg or fork risks. In short, rollup tx batches would be irreversible in seconds instead of a quarter hour.

2. Reducing consensus overhead via SNARKification: Beam proposes to “SNARK-ify” the state-transition function so every L1 block is accompanied by a succinct zk SNARK –  a prereq for synchronous, programmable execution sharding down the line. Validators will be able to verify a block and aggregate BLS signatures (and later post-quantum signatures) without processing every transaction, dramatically reducing the computational cost of consensus (and thus also reducing validator hardware requirements).

All of this points to a future Ethereum base layer that is more scalable and more decentralized. In particular, based rollups benefit the most from these consensus upgrades because the L1 will be better suited for handling their sequencing. By ordering transactions on L1, the MEV from based (and native based) rollups naturally flows to Ethereum block proposers and can be burned, re-centering more value accrual to ETH instead of centralized sequencers.

DA throughput is essential for rollup scaling, particularly for future performance rollups that require throughput processing upwards of 100K TPS. Ethereum proto-danksharding (Dencun + Pectra upgrades) have increased the target and maximum number of blobs per block to 6 and 9, respectively, increasing blob data capacity to 8.15 GB/day (~94 KB/s; ~1.15 MB/block). But that’s not near enough. By 2030, Ethereum could achieve full danksharding, which targets 64 blobs (1 blob = 128 KB) per block (~8 MB per ~4 s slot = 2 MB/s).

While this is an order of magnitude improvement, it’s still a far way from the ~20 MB/s needed for a performance rollup like MegaETH. But, Ethereum’s roadmap goes further: Data Availability Sampling (DAS) via schemes like PeerDAS (ETA: H2 2025 / H1 2026) will let nodes verify availability of data without downloading it all and bring the blob target per block to 48+ via data sharding. With ideal danksharding and DAS, Ethereum could support 16 MB per 12s slots, which equates to about 7,400 simple tx/s – and as high as 58,000 TPS with compression gains (e.g., aggregating signatures, compressing addresses, etc.) and even higher with by incorporating Plasma or Validium (where only state roots, not full data, are onchain). While there are security vs. scalability tradeoffs to consider with offchain scaling (e.g., operators disappear or freeze funds), the point is that by 2030, Ethereum could provide a spectrum of DA options enshrined within its protocol: fully guaranteed onchain data for rollups that need security over scale, and the ability to opt into external DA for those that need scale over throughput. 

Summing up, Ethereum’s DA upgrades make Ethereum increasingly more suitable for rollups. The caveat is that Ethereum’s current throughput is far from being able to support high-throughput sectors like payments, social, or gaming. Even assuming simple ERC-20 transfers requiring ~200 bytes of blob data, the back-of-the-envelope calculation is ~20 MB/s of raw DA bandwidth. More complex transactions produce larger state diffs (e.g., Uniswap swaps would triple that figure to ~60 MB/s!). Full danksharding alone won’t get anywhere near this bandwidth, so scaling throughput would depend on a clever combination of data compression and offchain scaling. In the meantime, performance rollups will need to use alternative DA solutions like Eigen DA that can support them with ~15 MB/s data throughput currently and a roadmap to 1 GB/s. Newer solutions like Hyve promise 1 GB/s modular DA with sub-second availability. These are the kind of DA solutions that can give web3 apps web2-like speed and UX.

“Ethereum aims to be the world ledger: the platform that stores civilization's assets and records, the base layer for finance, governance, high-value data authentication, and more. This requires two things: scalability and resilience.” – Vitalik

Ethereum 2030 will be much more ready to take on this role with its core protocol and rollup-centric upgrades. As I’ve outlined here, these upgrades across the stack will support two rollup archetypes: one gravitates towards being “more Ethereum” for security and credible neutrality and the other toward being “less Ethereum” for extreme throughput and economic independence. Ethereum’s roadmap isn’t forcing any single path, but rather providing the grounds to be flexible enough for both to thrive:

• Aligned rollups ensure that the most valuable, interconnected applications remain strongly secured by Ethereum. With based rollups, you get Ethereum-level liveness because the L1 validator that produces a rollup block also sequences that block; with native rollups, you get Ethereum-level execution security because every rollup state transition is re-executed and proven inside the L1; and a native based rollup (or “ultrasound rollup”; aka execution shard) you get both 100% execution security + 100% liveness – essentially making the rollup a shard of Ethereum L1 itself. These rollups will benefit Ethereum L1’s value accrual: MEV from based rollups flow directly to Ethereum validators and, with MEV burn, can increase ETH scarcity; verifying proofs of native rollups by calling the EXECUTE precompile will cost gas, creating a new revenue stream for ETH. In a world where most of DeFi and institutional finance runs on a handful of aligned rollups, ETH would capture fees from that entire economy. And the combination of Ethereum’s censorship resistance + MEV value capture are two of the most important things for Ethereum to fulfill its mission to become the world ledger.

• Performance rollups enable Ethereum’s ecosystem to encompass the full spectrum of blockchain use cases, including those requiring massive scale. These chains will most likely act as a bridge to mainstream adoption. While they might introduce (semi-)trusted elements, they keep Ethereum at least as the final settlement layer and interoperability hub. The presence of performance + aligned rollups means the Ethereum ecosystem can credibly claim to support both top-tier security and top-tier throughput applications. L2 heterogeneity + interoperability is good for Ethereum. While these rollups may be less economically aligned with ETH, they can still drive net-new demand for ETH by using ETH as a gas token, medium of exchange, DeFi denomination, and as a core asset in novel apps that can only exist in these high-capacity environments. Notably, the DA discussion above that suggests Ethereum DA could plausibly support 100K+ TPS means that even performance chains may eventually return to Ethereum for DA instead of depending on a modular alternative (e.g., for social alignment, credible neutrality, stack simplification, etc.). Of course, they will always have the choice to use an alternative DA to save on costs or boost performance if needed, but the point is that Ethereum DA, data compression, and offchain data management will all progress to make Ethereum L1 more competitive.

The main exception I see are rollups with strong affiliation with trusted companies (e.g., Coinbase’s Base, Robinhood Chain), where trust in these companies > trust in trustless systems (an effect particularly pronounced in new and non-technical users). In such cases, the credibility and accountability of the affiliated company becomes the primary assurance mechanism. As such, they could forgo some Ethereum alignment while staying competitive because users are willing to “trust the brand” like they do in web2. However, a large part of their adoption would depend on B2B trust, where, for instance, JPMorgan Chain trusts Robinhood Chain over arguably stronger assurances on Ethereum and aligned rollups.

Otherwise, the gradual consolidation away from the middle ground is likely a natural consequence of the maturation of these two tracks. The reasoning is straightforward: those in the middle would neither be as aligned nor as performant. Users concerned with security and composability would prefer rollups that are more aligned with Ethereum, while users who care more about cheap, fast transactions will gravitate to the platforms that deliver that best. On top of that, performance gains on aligned rollups will keep improving with advancements in preconfirmations, faster slots, and faster finality on L1, such that there isn’t much need for a moderate performance gain. In general, the former would be better suited for institutional DeFi and the latter for retail consumer apps.

The resources to run a successful rollup (from attracting liquidity to maintaining infrastructure) are significant, and by 2030 we will likely see more consolidation where stronger networks absorb the communities of weaker ones. We’re already seeing the early signs. In the long run, an ecosystem of a few major hubs with clear value propositions will beat hundreds of undifferentiated ones.

Special thanks to mteam, Patrick, Amir, Jason, Jünger, and Bread for the useful discussions and feedback!

Disclaimer

This article is prepared for general information purposes only. This post reflects the current views of its authors only and is not made on behalf of Lemniscap or its affiliates and does not necessarily reflect the opinions of Lemniscap, its affiliates or individuals. The opinions herein may be subject to change without this article being updated.

This article does not constitute investment advice, legal or regulatory advice, investment recommendation, or any solicitation to buy, sell or make any investment. This post should not be used to evaluate the making of any investment decision, and should not be relied upon for legal, compliance, regulatory or other advice, investment recommendations, tax advice or any other similar matters.

All liability in connection with this article, its content, and any related services and products and your use thereof, including, without limitation, the implied warranties of merchantability, fitness for a particular purpose, and non-infringement is disclaimed. No warranty, endorsement, guarantee, assumption of responsibility or similar is made in respect of any product, service, protocol described herein.