Eclipse Labs releases a reflective article and leadership plan

The founder of Eclipse, Neel Somani, stated that a year of operating dozens of testnet chains has led the team to realize that the future of blockchain scalability may lie in shared infrastructure rather than thousands of application-specific Rollups.

In a blog post titled "A Year Around the Sun With Eclipse," Somani elaborated on these lessons, which prompted Eclipse to shift from its initial Rollup-as-a-Service model to the Eclipse Mainnet—a general-purpose Ethereum Layer 2 powered by the Solana Virtual Machine (SVM).

Somani also mentioned that as the company enters its next phase, Eclipse is preparing for a leadership transition. He plans to step down as CEO and take on the role of chairman in early 2024, with a trusted leader from within Eclipse succeeding him as CEO. Somani stated that he will continue to be closely involved in the company's strategic direction.

Meanwhile, Somani cautioned against overstating Eclipse's progress. The company still needs to launch the Eclipse Mainnet, build an active ecosystem, achieve proven decentralization, and further improve its bridging architecture.

These changes occurred after a broader reflection on Eclipse's original vision.

The company initially collaborated with application developers to deploy customizable SVM-based Rollups and ultimately operated over 30 testnet chains with participating teams. However, this experience exposed what Somani described as significant operational and economic challenges.

Maintaining separate chains requires infrastructure that includes sorters, full nodes, indexers, and engineering support, while the team must also manage integration, bridging, and other operational demands.

According to Somani, commercial Rollup-as-a-Service providers could charge around $60,000 to $100,000 per year, plus a share of sorter fees.

From Custom Chains to Shared L2

Eclipse's experience also challenges one of the core arguments for application-specific blockchains: customization.

While having complete control over the blockchain architecture may seem appealing, Somani believes that most applications do not require extensive protocol layer customization. Unique modifications can introduce additional auditing requirements, technical complexity, and interoperability challenges.

There are exceptions, however. Projects that prioritize sovereignty, ownership, or direct control over upgrades and forks may still benefit from operating dedicated chains.

But Eclipse concluded that the broader market might be better served by fewer high-performance shared networks.

This thinking ultimately shaped the Eclipse Mainnet.

Eclipse Chooses Ethereum as the Settlement Layer

Eclipse also abandoned its previous plan to build its own settlement layer.

The initial architecture envisioned an Eclipse settlement network running application-specific Eclipse Rollups. A custom settlement layer could have provided greater flexibility and lower potential costs.

However, Eclipse ultimately determined that Ethereum already offers many of the features it hoped to gain from a settlement layer, including liquidity, security, verifiability, and neutrality.

Betting on a Different Type of Modular Blockchain

Eclipse Mainnet is the culmination of these lessons.

Eclipse no longer provides infrastructure for potentially thousands of isolated application Rollups but instead bets on a general-purpose Layer 2 where applications can share execution infrastructure and liquidity while retaining Ethereum as the settlement layer.

This architecture is also an unusual attempt to merge the design philosophies of Ethereum and Solana: Ethereum provides the settlement foundation, while SVM aims to deliver high-performance execution.

More broadly, Eclipse's directional shift indicates that once infrastructure moves from theory to practical operation, assumptions about blockchain scalability may change.