————————————————————
Next-generation L1 projects like Hyperliquid, Monad, and Sonic have somewhat alleviated their dependence on L2, but they now face new challenges.
📍This article will take you through eight sections to deeply analyze how these three projects address the blockchain trilemma and innovate consensus mechanism design.
1) Can consensus mechanisms save blockchain?
2) Hyperliquid: A high-speed, low-cost decentralized trading platform
3) Monad: An L1 optimized for performance and scalability
4) Sonic: High-speed, low-cost EVM-compatible L1
5) Chart comparison
6) Leveraging the Ethereum ecosystem
7) Is L2 Ethereum's scalability answer?
8) Conclusion and thoughts
1) Can consensus mechanisms save blockchain?🔻
Consensus mechanisms are the 'rules of the game' for blockchains, ensuring every computer in the network agrees on which transactions are valid and are verified and recorded in order on the chain. Simply put, they keep distributed networks consistent and prevent cheating.
Blockchain faces a classic dilemma—the blockchain trilemma: speed, security, and decentralization. Projects typically can only optimize two of these, sacrificing the third.
🔹Speed: Transactions must be fast for a good user experience.
🔹Security: Preventing hackers from tampering with data or double spending (spending the same money twice).
🔹Decentralization: Power is dispersed, not controlled by a few individuals.
Consensus mechanisms are the 'firewall' that ensures the network operates in sync, with each node processing transactions in the same order. Here are several mainstream consensus mechanisms.
——PoW: Miners solve mathematical problems using computing power to add blocks and receive cryptocurrency rewards. Secure but energy-intensive and slow (e.g., Bitcoin, Ethereum before 2022).
——PoS: Validators stake tokens to compete for block creation opportunities, which is energy-efficient and faster, but may favor the 'wealthy' (like Ethereum and Cardano after 2022).
——DPoS: Token holders vote to elect representatives to verify transactions, fast and scalable, but may lack decentralization (e.g., EOS, Tron).
——PoA: Verified by trusted nodes, efficient but low decentralization (e.g., VeChain).
Although blockchain promises decentralization, actual performance often falls short, especially for blue-chip projects.
🔹BTC: An average of 7 transactions per second (TPS).
🔹ETH (after PoS): 15-30 TPS.
🔹Comparison to Visa: Average 1700 TPS.
This gap leads to network congestion, delays, and high fees, exposing scalability bottlenecks.
🔸1.2 Next-generation consensus mechanisms
Emerging Layer-1 blockchains, such as Hyperliquid, Monad, and Sonic, are designing new consensus mechanisms specifically to address the challenges of speed, scalability, and impact, while enhancing user trust.
This article will delve into how these three projects tackle the blockchain trilemma and innovate consensus mechanism design. We will discuss the project background, consensus mechanisms, relationships with Ethereum, scalability solutions, practical applications, funding and governance methods, as well as the challenges they face one by one.
2) Hyperliquid: A high-speed, low-cost decentralized trading platform🔻
Hyperliquid is an L1 blockchain designed for high-speed, low-cost decentralized trading, built on two main pillars:
1️⃣HyperCore: An on-chain engine supporting perpetual futures and spot order books, with block confirmations taking only one second.
2️⃣HyperEVM: An Ethereum-compatible smart contract platform.
Traditional L1s always have to compromise between decentralization, performance, and ease of use, while Hyperliquid aims to create a high-performance, comprehensive trading ecosystem. HyperCore can handle up to 200,000 orders per second, and future upgrades through node software can further enhance performance.
HyperEVM allows Hyperliquid to tap into Ethereum's smart contract ecosystem, opening HyperCore's liquidity and financial tools to developers, assisting DApps in seamless interaction with blockchain components while balancing efficiency and user experience.
🔸2.1 Consensus mechanism: HyperBFT
Hyperliquid initially used the Tendermint consensus algorithm, but to support high-frequency trading and higher throughput, developed HyperBFT. This is a hybrid mechanism combining PoS and BFT based on the HotStuff protocol, optimizing for high throughput, low latency, and strong security.
PoS: Validators stake $HYPE tokens to generate blocks, energy-efficient and secure.
HyperBFT: More efficient than traditional PoW while maintaining robust defenses.
🔸2.2 Scalability and speed
HyperBFT's median block confirmation time is only 0.2 seconds, with delays under 0.9 seconds. The on-chain order book rivals centralized exchanges, supporting 50x leverage, one-click trading, and stop-loss features.
Hyperliquid performs exceptionally in high throughput scenarios, with a single node capable of handling 200,000 TPS without sharding. Current limitations mainly stem from network latency and validator distribution.
🔸2.3 Challenges
Low number of validators (security): Hyperliquid currently has only 16 validators compared to Ethereum's 800,000+, making it appear quite centralized. Plans to increase validators in the future to enhance decentralization.
Unverified attack resistance: In 2024, a $2.3 billion USDC bridge was hacked, exposing the security risks brought by centralization.
Centralization controversy: In March 2025, the $JELLY token incident sparked heated discussions. A trader manipulated the platform's liquidation system, causing a 400% price surge, resulting in Hyperliquid's liquidity pool sustaining unrealized losses of $600,000 to $10 million. Ultimately, the platform intervened and delisted the $JELLY contract, leading to debates on decentralization and governance transparency.
High leverage risks: On March 13, 2025, a large trader liquidated a long position in $ETH using high leverage, causing approximately $4 million in losses for the HLP liquidity pool, highlighting vulnerabilities in market manipulation and risk management.
Competitive pressure: Hyperliquid's closed-source code and lack of automatic validator penalty mechanisms limit transparency. Fierce competition with high-throughput platforms like Solana, Monad, MegaETH, and dYdX.
Scalability bottlenecks: Despite designs supporting high TPS, extreme scenarios (such as large-scale leveraged trading) may face liquidity pressures or validator coordination delays.
3) Monad: An L1 optimized for performance and scalability🔻
Monad is an Ethereum Virtual Machine (EVM) compatible L1 blockchain focused on scalability and performance, employing parallel execution and the MonadBFT consensus mechanism.
Monad aims to achieve 10,000 TPS, with a block generation time of 0.5 seconds and a confirmation time of 1 second, while maintaining decentralization and addressing Ethereum's bottlenecks (such as slow speeds, high fees, and insufficient scalability). Its testnet went live on February 19, 2025, with the mainnet expected to launch by the end of this year.
🔸3.1 Consensus mechanism: MonadBFT
The core of Monad is MonadBFT, an optimized version of the HotStuff BFT protocol, combining pipelined execution and efficient communication.
MonadBFT: Simplifies HotStuff's three-phase process to two phases, improving validator speed. Validators take turns being leaders, proposing blocks and collecting votes to form a 'quorum certificate' (QC), proving that the previous block has reached consensus. A timeout mechanism ensures that the network can still operate if the leader fails.
Parallel execution: Nodes first reach consensus on the order of transactions, then execute transactions in parallel using multithreading, employing optimistic execution strategies to ensure results are consistent with sequential execution while significantly increasing throughput.
PoS: Validators participate by staking tokens, with economic incentives ensuring network security.
MonadBFT provides scalable and reliable confirmations for real-time dApps by reducing communication overhead.
🔸3.2 Scalability and speed
Monad combines MonadBFT and parallel execution, processing transactions without sharding, theoretically exceeding 10,000 TPS with confirmation times under 1 second. Actual performance depends on network latency and validator distribution.
🔸3.3 Challenges
Execution complexity: Optimistic parallel execution may lead to inconsistencies, rollbacks, or vulnerabilities, increasing development and maintenance costs.
Network latency: Actual TPS and confirmation times are affected by validator distribution and latency, which may not meet expectations.
Unverified scale: Before the mainnet launch, the promise of 10,000 TPS has not been fully verified, and there may be bugs or bottlenecks.
Competitive pressure: High-throughput platforms like Sonic, Arbitrum, and Solana may compete for developers and users.
Learning curve: Despite being EVM-compatible, unique systems like MonadBFT and MonadDB may slow developers' onboarding.
Centralization risks: Initially controlled by the foundation, the token model is centralized, which could threaten long-term decentralization.
4) Sonic: High-speed, low-cost EVM-compatible L1🔻
Sonic is an EVM-compatible L1 blockchain inheriting from the Fantom Opera ecosystem, focusing on high throughput and sub-second transaction confirmations.
Sonic's latest consensus protocol SonicCS 2.0 brings significant improvements: consensus speed doubles, and memory usage per round is reduced by 68% (from 420MB to 135MB), lowering validator resource requirements and enhancing scalability.
These upgrades address the following issues:
——Slow transaction processing
——High operational costs
——Ecosystem fragmentation
Sonic incentivizes dApp development and adoption through its FeeM program (returning up to 90% of network transaction fees to developers).
🔸4.1 Consensus mechanism: Lachesis
Sonic's Lachesis consensus combines Directed Acyclic Graph (DAG) and Asynchronous Byzantine Fault Tolerance (ABFT), surpassing the foundation of Fantom Opera.
——ABFT: Validators asynchronously process transactions and exchange blocks, eliminating the sequential delays of traditional PBFT systems, increasing throughput and resilience.
——DAG: Transactions are represented by vertices, with dependencies as DAG edges, supporting concurrent block addition, forming a mesh structure rather than a single chain, accelerating verification.
——PoS: Validators must stake at least 500,000 $S tokens, processing transactions in batches to form event blocks, which become the main chain 'root' after enough validators confirm, achieving sub-second confirmations.
🔸4.2 SonicCS 2.0: Consensus mechanism upgrade
Sonic launched SonicCS 2.0 on March 27, 2025, using DAG + overlapping elections, reducing computation and memory usage by 68%. Experiments show that Sonic's mainnet data averaged a 2.04x speedup over 200 rounds (from 1.37x to 2.62x), supporting over 10,000 TPS with sub-second confirmation times. SonicCS 2.0 is about to go live on the mainnet, with a detailed technical report to be released.
🔸4.3 Scalability and speed
Lachesis consensus combines the flexibility of DAG and the integrity of ABFT, achieving fast and secure transaction confirmations without sharding, supporting seamless scalability. SonicCS 2.0 theoretically can reach 396,000 TPS, with actual tests achieving 5,140 TPS (data source: @AndreCronjetech)
）。
Sonic is fully EVM-compatible, optimizing performance through vectorized operations and overlapping elections, enhancing validator and dApp efficiency.
🔸4.4 Challenges
Consensus complexity: Under high load, Lachesis may trigger complex dependencies or verification delays, increasing vulnerability risks.
Developer adaptation: Despite being EVM-compatible, features like vectorized voting in SonicCS 2.0 may require developers to adjust their workflows, slowing adoption.
Network latency: Sub-second confirmations and 10,000 TPS depend on validator distribution and latency, and actual performance may decline.
Unverified scale: Before SonicCS 2.0 goes live, the promise of 10,000 TPS has not been fully verified, potentially with bottlenecks or bugs.
L2 competition: Ethereum L2s (like Optimism, zkSync) offer similar performance at lower costs. Sonic needs to enhance interoperability through the Sonic Gateway bridge, but as an independent L1, it still faces challenges.
Centralization risks: The 500,000 $S staking threshold and early control by the Sonic Foundation could lead to power concentration, affecting decentralization.
5) Chart comparison🔻
6) Leveraging the Ethereum ecosystem🔻
Hyperliquid, Monad, and Sonic are all EVM-compatible, allowing developers to use familiar tools and smart contracts to deploy dApps on high-performance infrastructure. This brings low-cost, high-throughput transactions while leveraging Ethereum's ecosystem without rewriting code.
Empowering diversified dApps, these L1s provide sub-second confirmations and high TPS, suitable for various dApps:
🔹Hyperliquid: Provides fast, secure DEX trading, with on-chain order books rivaling centralized exchanges.
🔹Sonic: Sub-second confirmations support efficient DeFi applications.
🔹Monad: 10,000 TPS, 1-second block times, and single-slot confirmations, comprehensively enhancing performance.
Beyond Web3: Enterprise potential
The speed and scalability of these networks make them suitable for enterprise scenarios such as finance, supply chains, and payments. Retailers can handle high-frequency payments to reduce costs, and healthcare institutions can protect patient data in real-time while remaining compatible with existing systems.
7) Is L2 Ethereum's scalability answer?🔻
Why do we need new L1s like Hyperliquid, Monad, and Sonic? Hasn't L2 already solved the scalability problem?
L2 (such as Arbitrum, Optimism, Base) enhances L1 scalability by processing transactions off-chain. Arbitrum can achieve 4000 TPS, and Base plans to implement 0.2 second Flashblocks by mid-2025. However, L2 relies on Ethereum's security and confirmation mechanisms, inheriting its limitations. For example, optimistic Rollups require fraud proofs, which may cause confirmation delays, affecting the user experience of applications that require fast confirmations.
New L1s directly address these issues through advanced consensus mechanisms without relying on Ethereum's infrastructure, avoiding bottlenecks from fraud proofs or L1 block times. However, new L1s also carry risks, such as potentially lower degrees of decentralization, higher development costs, and a lack of Ethereum's historical trust and stability.
Some argue that 'L2 is faster, cheaper, and safer, while L1's complex consensus mechanisms introduce risks and high costs.' However, L2's scalability is limited by Ethereum, and transaction confirmation times depend on L1 block confirmations. New L1s pursue independence and speed but must prove they can securely serve hundreds of millions of users.
Core issue: Can new consensus mechanisms completely resolve L1 scalability issues, or must they be combined with L2 compromise solutions? This requires ongoing research and discussion within the blockchain community.
8) Conclusion and thoughts🔻
Current market liquidity is fragmented and unstable, making it a significant challenge to attract the attention of users and developers. To drive adoption, the actual needs of users and developers must be prioritized.
In simple terms, users do not care how cool the underlying technology is; they want:
——Smooth experience: Fast transactions, low fees, especially for micro-transactions.
——Security guarantees: Assets and data are as stable as a mountain, enhancing trust.
——Rich features: There are enough fun or practical applications on-chain.
Both L1 and L2 need to fight for these demands. Instead of blindly pursuing the 'strongest technology' or overly optimizing consensus mechanisms, it is better to be pragmatic and build the most user-friendly network for developers and users.
Overall, new L1s like Hyperliquid, Monad, and Sonic break free from L2 dependency through innovative consensus mechanisms but also face challenges. For example, Hyperliquid has a low number of validators (only 4 nodes could lead to collusion risks), exposing security vulnerabilities. Increasing validators, enhancing bridging security, and implementing real-time monitoring and anomaly detection can improve resilience.
Developers need to find a balance between security, scalability, and decentralization, establishing trust through proactive risk management. Users should pay attention to the platform's security measures, and developers must prioritize building robust defense mechanisms.
Let 'developers handle the technical work', optimizing the trade-offs of consensus mechanisms; while also not forgetting the users, who just want responsive, efficient, safe, and decentralized applications. These new designs are pushing the limits of consensus mechanisms, and how they will evolve and integrate with competitors in the future.
🔹Original compilation link: https://x.com/castle_labs/status/1912153794571272637?t=aom90N27n8ub4S_x3eR5Qw&s=19