Puffer UniFi (Based Rollup) compared with mainstream Rollup

Author: Chen Mo Source: X, cmDecentralized Finance

Key Point: Puffer is utilizing Restaking to build a more efficient Based Rollup solution, Decentralization Sorter, and liberate the inter-chain liquidity interaction. Based Rollup will form a different Ethereum ecosystem pattern from mainstream Rollups.

• Puffer LRT: ETH2.0 staking service based on Eigenlayer.
• Puffer UniFi: Based Rollup enhancement solution with pre-commitment technology introduction.
• Puffer UniFi AVS: Use UniFi AVS to build highly synchronized AppChain ecosystem and reduce Liquidity fragmentation.
• The difference between Based Rollup and mainstream Rollup solutions comes from continuous trade-offs in decentralization, efficiency, benefits, and other aspects to adapt to different types of application scenarios.

Research Report

1/6 · Basic Concepts of Rollup

Rollup is a technology used to scale blockchains like ETH and other, with the aim of enabling the blockchain to process more transactions while maintaining lower fees. In simple terms, it ‘bundles’ a large number of transactions together and then processes these transactions off-chain (as opposed to the Mainnet), shifting the computing and storage work outside the ETH Mainnet (L1). This way, the burden on the Mainnet is reduced, but the final results of the transactions will still be submitted to and verified on the ETH Mainnet, thereby maintaining the security and immutability of the data.

Rollup is divided into two main types:

• Optimistic Rollup: Assuming that most people are honest, all transactions are assumed to be correct by default, and validation only occurs when someone questions. This method is fast, but has a certain challenge period.
• ZK Rollup (Zero-Knowledge Proof): Use mathematical proofs to ensure the correctness of each transaction, ensuring their legality when packaging transactions, so there is no need for additional challenge periods. This method is more secure, but the technology is more complex.

In general, the core idea of Rollup is to “do a lot of work off-chain and only put the results on-chain”, so as to make the processing speed of the blockchain faster and the fees lower.

So this process can be simply broken down into two parts:

• Bulk Packaging - Increase volume/throughput
• Transaction Confirmation - Speed up confirmation

2/6 · Based Rollup

ETH Foundation researcher Justin Drake proposed the concept of Based Rollup in March 2023, which mainly targets a crucial role between L1 and L2 - the sequencer.

The sequencer is responsible for the transaction sequencing and packaging from L2 to L1, with its business scope covering batch packaging and transaction confirmation mentioned above. However, the current sequencing solutions used by most L2 are centralized solutions controlled by a single or a few entities. In the early stages of industry development, this may not show significant issues. However, as the funds on each chain grow and the infrastructure becomes more mature, centralized sequencers will become a problem that “needs to be solved sooner or later”.

Currently, there are two proposed solutions:

• Decentralized sorter based on L2
• Transfer to Based Rollup on L1

The main representative of the Decentralization sorter based on L2 is Metis, and the implementation method is to establish a new powerful Consensus to make the sorter network more robust.

The solution of Based Rollup is that it removes the separate sorter network mechanism design and transfers the responsibility of transaction sorting from L2 to L1. The proposer of the block, which is validated by the ETH node on L1, is responsible for the transaction sorting. This design does not require the reconstruction of Consensus, fully utilizes the existing ETH node resources, and avoids the risks of centralized sorter, grafting the security level of the sorter onto the ETH Mainnet.

However, the problems faced by the Based Rollup are also limited by the classic trilemma. Since it has obtained the security of the Mainnet, it also means that it must rely on the performance of the Mainnet (with a block time of about 12s), which will greatly lag behind the transaction confirmation time of conventional L2 sorter solutions.

3/6 · Puffer UniFi Optimization Based Rollup Solution Analysis

To break through this trilemma, it is necessary to introduce new roles or mechanisms to achieve balance. UniFi proposed by Puffer aims to introduce “pre-confirmation” to solve this problem.

In existing Rollup solutions, users usually obtain “soft confirmations” from centralized sequencers. These sequencers can assure users that the transaction will be included in the blockchain after the user submits it. This kind of soft confirmation essentially comes from the trust in this centralized power. Once this power becomes corrupt, transaction sequencing/confirmation cannot be fairly guaranteed.

Puffer UniFi is confirmed to have the following two features:

1. Validators of the ETH network provide assurance: Before the Block is packed, the validators of the ETH mainnet make a commitment in advance to the pre-confirmed Rollup transactions - to include these transactions in the Block they are about to propose.
2. Punishment mechanism: If validators fail to fulfill their pre-committed promises, i.e. fail to include the pre-committed transactions in the Block, they will face penalties.

So UniFi bypasses the problem of being restricted by the Mainnet block time in the Based Rollup mechanism through “pre-confirmation”, which means that in the process of transaction confirmation, the ETH mainnet nodes are made to bear “additional work” to ensure the efficiency of transaction confirmation (~100ms).

With the pre-confirmation mechanism, it can greatly improve the efficiency of Mainnet transaction confirmation. So, is L2 still needed to improve Mainnet performance? “quick confirmation of transactions”. This is an expansion in both horizontal and vertical dimensions. Pre-confirmation solves the problem of fast and fair transaction confirmation, while batch packaging of transactions still needs to be completed through Layer 2 or off-chain, and then the packaged transactions are sent to the Mainnet and confirmed through the sorter.

4/6 · Puffer UniFi AVS

Ether Node assumes ‘additional work’ in the ‘pre-commit’ mechanism, how should this additional work and the corresponding penalty mechanism be implemented?

Puffer proposes UniFi AVS, which is a solution based on EigenLayer without introducing additional mechanisms and Collateral, but utilizing its ETH ‘Restaking’ to package UniFi into an AVS service. Any validators who stake again on EigenLayer can participate in UniFi AVS to provide pre-confirmation services and design corresponding penalty mechanisms in AVS.

Overall, with Restaking and AVS, “pre-confirmation” is more smoothly implemented, which is currently one of the most fitting and practical AVS solutions, adding to the Based Rollup solution, and it was not easy to achieve this functionality before.

5/6 · Core Value

Using UniFi AVS to build an application chain will become an upgrade solution for Based Rollup, compared to other Rollups:

• Decentralization ranking based on L1
• Integrated pre-confirmation technology improves transaction confirmation speed
• Create earnings for Restaking
• Synchronous Composability

Synchronized composability is one of the core features UniFi uses to solve the fragmentation problem. In the current multi-Rollup environment of Ethereum, different Rollups run independently of each other, leading to Liquidity and user dispersion, and the interaction of assets and contracts between different Rollups needs to be completed through cross-chain bridges and other tools, which not only increases complexity but also brings high costs and security risks.

Because UniFi is based on L1’s Decentralization sorting, the ETH Mainnet Node acts as the ‘General Sorter’, so it allows interaction between different Rollups (based on UniFi) within the same Block without the need for cross-chain bridges. In this way, users can seamlessly operate between different Rollups and AppChains within the UniFi ecosystem. By doing so, UniFi provides users with a smooth experience similar to a single chain, effectively reducing the fragmentation of Liquidity. This is similar to the ultimate goal of chain abstraction, and its advantage lies in a more native and simple implementation, without the need to introduce additional Consensus and restrictions (such as establishing a new Consensus layer to aggregate Liquidity and user demands).

6/6 · Summary

Overall, the entire UniFi business depends on whether Based Rollup is a necessary Rollup solution.

From the current L2 state, most of the solutions are actually more centralized than users’ impressions, but this centralization issue will not affect user interests in the short term. Similarly, the resolution of this issue in the short term will not bring wealth effects because it is far from the user end. The reason why this issue is not obvious is that no application is directly affected by the centralization degree of the sorter. Just like the inscription period of the BTC ecosystem, all the stories of getting rich are completed under a highly centralized indexer background.

However, if we take a long-term perspective, when discussing the Decentralization sorter, while considering Liquidity and fragmented experience, this is an inevitable thing, so let’s bring in some imagery for a comparison:

Based Rollup is a solution that is currently more closely related to the Ethereum Mainnet, it is more like an ‘affiliated’ Rollup of the Ethereum Mainnet, or described as the close guard of the Ethereum, it can be infantry, cavalry or artillery, and the communication between the barracks is directly managed by the Ethereum, and follows the fairest system of Ethereum, without the need for a third party to intervene, there is no risk of information transmission errors and intentional misconduct.

Currently, the mainstream Optimistic Rollup is a more “decentralized” Rollup for Ethereum, like the autonomy of feudal lords, they are diverse, building their favorite cities and systems. Whether this system is fair or not depends on the feudal lords themselves. At the same time, communication between the feudal lords requires an additional department for scheduling and regulation, and this department is the “chain abstraction.” To control these feudal lords, this department of chain abstraction must be powerful enough, not only to have the ability to control the feudal lords but also to ensure that it is not corrupted itself.

The controversy between these two solutions lies in whether Rollup should return value to Ethereum or should be independently autonomous, closer to Ethereum for greater security advantages, saving a lot of the cost of Decentralization, and having natural synchronous composability. Autonomy means that operators can capture great benefits, but at the same time, they have to spend a lot of costs to build security based on Decentralization and rely on externally provided composability.

The inclusiveness and freedom of blockchain determine that these two solutions will coexist. The source is the continuous trade-off of various solutions in terms of decentralization, efficiency, benefits, etc., to adapt to different types of application scenarios. This exploration will continue, and only continuous technical adjustments can create a new future for developers.