Parallel EVM Research Summary: Ecosystem and Features Explained

Intermediate8/27/2024, 1:20:53 AM
This article explores Parallel EVM technology, and its applications and potential in blockchain. It has attracted attention and investment from leading venture capital firms. Parallel EVM supports more complex and efficient blockchain applications through comprehensive performance optimization, achieving differentiation in the open-source ecosystem while balancing decentralization and high-performance needs. The article analyzes the technical challenges of Parallel EVM, such as data consistency, state access efficiency, and transaction conflict detection, and introduces unique solutions from projects like MegaETH and Artela.

TL;DR

  1. Parallel EVM has attracted the attention of top venture capital, and many projects have begun to explore this direction, such as Monad, MegaETH, Artela, BNB, Sei Labs, Polygon, etc.
  2. Parallel EVM is not just about achieving parallel processing but also involves comprehensive performance optimization of each component of the EVM. Through these optimizations, more complex and efficient blockchain applications are supported.
  3. Parallel EVMs must stand out in the open-source ecosystem by balancing decentralization and high performance, while also addressing potential security issues and market acceptance challenges. The complexity of multi-threaded programming introduces the challenge of managing multiple transactions at once, requiring effective solutions to ensure system stability and security.
  4. In the future, parallel EVM will promote the implementation of on-chain central limit order books (CLOB) and programmable central limit order books (pCLOB), which will greatly enhance the efficiency of DeFi activities, and the DeFi ecosystem is expected to grow significantly.
  5. Integrating other high-performance virtual machines (AltVM) into the Ethereum ecosystem will greatly enhance both performance and security. This approach leverages the strengths of each virtual machine, driving further development of Ethereum.

This year, Parallel EVM has garnered attention from leading venture capital firms like Paradigm and Dragonfly, capturing significant market interest. Unlike the traditional EVM, which processes transactions sequentially and can cause congestion and delays during peak periods, Parallel EVM leverages parallel processing technology to execute multiple transactions simultaneously, dramatically speeding up transaction processing. As complex applications such as on-chain games and account abstraction wallets become more prevalent, the demand for blockchain performance grows. To accommodate a larger user base, blockchain networks must handle high transaction volumes efficiently. Consequently, Parallel EVM is vital for the advancement of Web3 applications.

However, implementing Parallel EVM comes with common challenges that require precise technical solutions to ensure stable system operation.

  • Data Consistency: In Parallel EVM, multiple transactions occurring simultaneously may require reading or modifying account information concurrently. Effective locking mechanisms or transaction processing methods are needed to ensure data consistency during state modifications.
  • State Access Efficiency: Parallel EVM must quickly access and update states, demanding efficient state storage and retrieval mechanisms. Optimizing storage structures and access paths, such as using advanced data indexing techniques and caching strategies, can significantly reduce data access latency and enhance overall system performance.
  • Transaction Conflict Detection: In parallel execution, multiple transactions may depend on the same data state, making transaction ordering and dependency management complex. Complex scheduling algorithms are required to identify and manage dependencies between parallel transactions, detect potential conflicts, and decide on handling methods to ensure parallel execution results are consistent with serial execution.

For example, MegaETH decouples transaction execution tasks from full nodes, assigning different tasks to specialized nodes to optimize overall system performance. Artela uses predictive optimistic execution and asynchronous preloading technologies to analyze transaction dependencies with AI and preload required transaction states into memory, improving state access efficiency. BNB Chain developed specialized conflict detectors and re-execution mechanisms to enhance transaction dependency management, reducing unnecessary re-executions, etc.

To understand the development direction of Parallel EVM in depth, here are nine selected high-quality articles on the topic, providing comprehensive perspectives on different chains’ implementation plans, ecosystem studies, and future prospects.

MegaETH: Unveiling the First Real-Time Blockchain

Author: MegaETH; Date: June 27, 2024

MegaETH is an EVM-compatible Layer 2 aiming to achieve near-Web2 server real-time performance. Its goal is to push Ethereum L2 performance to hardware limits, offering high transaction throughput, ample computing power, and millisecond response times. This allows developers to build and combine complex applications without performance constraints.

MegaETH enhances performance by separating transaction execution tasks from full nodes and introducing parallel processing technology. Its architecture consists of three main roles: Sequencer, Validator, and Full Node.

  • Sequencer: Responsible for ordering and executing user-submitted transactions. After executing transactions, Sequencers send state changes (state diffs) to Full Nodes via a peer-to-peer (p2p) network.
  • Full Node: Receives state diffs from Sequencers and directly applies these changes to update its local blockchain state, avoiding transaction re-execution. This significantly reduces computational resource consumption and improves overall system efficiency.
  • Validator: Uses stateless verification schemes to validate blocks, allowing simultaneous verification of multiple blocks. This further enhances verification efficiency and speed.

This specialized node design allows different types of nodes to set independent hardware requirements based on their functions. For instance, Sequencers need high-performance servers to handle a large volume of transactions, while Full Nodes and Validators can use relatively lower-spec hardware.

Presenting Artela Scalability Whitepaper - Parallel Execution Stack and Elastic Block Space

Author: Artela; Date: 2024.6.20

Artela significantly enhances blockchain parallel execution efficiency and overall performance through several key technologies:

  1. Parallel Execution: By predicting transaction dependencies and grouping transactions, it uses multiple CPU cores to parallel process, improving computational efficiency.
  2. Parallel Storage: Optimizes the storage layer to support parallel data processing, avoiding storage bottlenecks and enhancing overall system performance.
  3. Elastic Computing: Supports multiple computers working together, creating elastic computing nodes and block space, providing higher transaction throughput and predictable performance for dApps.

Specifically, Artela’s predictive optimistic execution uses AI to analyze dependencies between transactions and contracts, predicting potential conflicting transactions and grouping them to reduce conflicts and re-executions. The system dynamically accumulates and stores historical transaction state access information for predictive algorithms. Asynchronous preloading loads required transaction states into memory to avoid I/O bottlenecks during execution. Parallel storage improves Merkleization and I/O performance by separating state commitments from storage operations, managing parallel and non-parallel operations independently to further enhance parallel efficiency.

Additionally, Artela’s elastic computing builds elastic block space (EBS). Traditional blockchains share a single block space among all dApps, leading to resource competition among high-traffic dApps, causing unstable gas fees and unpredictable performance. Elastic block space provides dedicated and dynamically scalable block space for dApps, ensuring predictable performance. dApps can apply for exclusive block space as needed, and as block space increases, Validators can extend processing capabilities by adding elastic execution nodes, ensuring efficient resource utilization and adapting to different transaction volumes.

Road to High Performance: Parallel EVM for BNB Chain

Author: BNB Chain; Date: 2024.2.16

On the BNB Chain, the team has taken several steps to achieve Parallel EVM to enhance transaction processing capacity and scalability. Key developments include:

Parallel EVM v1.0:

  • Scheduler: Allocates transactions to different threads for parallel execution to optimize throughput.
  • Parallel Execution Engine: Utilizes parallel processing on dedicated threads to execute transactions independently, significantly reducing processing time.
  • Local State Database: Each thread maintains its own “thread-local” state database to efficiently record state access information during execution.
  • Conflict Detection and Re-execution: Ensures data integrity by detecting and managing transaction dependencies, re-executing transactions in case of conflicts to ensure accuracy.
  • State Submission Mechanism: Once transactions are executed, results are seamlessly submitted to the global state database to update the overall blockchain state.

Parallel EVM v2.0

Based on parallel EVM 1.0, the BNB chain community has introduced a series of innovations to improve performance:

  • Streaming Pipeline: Enhances execution efficiency, allowing smooth transaction processing in the parallel engine.
  • Universal Unconfirmed State Access: Optimizes state information access by allowing other transactions to temporarily use results from unconfirmed transactions, reducing inter-transaction wait times.
  • Conflict Detector 2.0: Improved conflict detection mechanism for better performance and accuracy, ensuring data integrity while reducing unnecessary re-executions.
  • Scheduler Enhancements: Utilizes static and dynamic scheduling strategies for more efficient workload allocation and resource optimization.
  • Memory Optimization: Significantly reduces memory usage through shared memory pools and lightweight copy techniques, further improving system performance.

Parallel EVM v3.0

After the performance improvements of parallel EVM 2.0, the BNB chain community actively developed parallel EVM 3.0 with the following goals:

  • Reduce or Eliminate Re-execution: Introduces a Hint-based scheduler using external hint providers to analyze transactions and predict potential state access conflicts. This helps better schedule transactions and reduce conflicts, minimizing re-execution needs.
  • Modularity: Decomposes code into independent modules for better maintainability and adaptation to different environments.
  • Codebase Refactoring: Aligns with the latest BSC/opBNB codebase to ensure compatibility and simplify integration.
  • Thorough Testing and Validation: Conducts extensive testing under various scenarios and workloads to ensure solution stability and reliability.

Sei’s Parallel Stack

Author: Sei; Date: 2024.3.13

Sei Labs has created an open-source framework called Parallel Stack, designed to build Layer 2 solutions that support parallel processing technology. The core advantage of Parallel Stack lies in its parallel processing capability, leveraging advancements in modern hardware to reduce transaction costs. This framework employs a modular design, allowing developers to add or modify functionality modules based on specific needs, thereby adapting to various application scenarios and performance requirements. Parallel Stack can seamlessly integrate with the existing Ethereum ecosystem, enabling applications and developers to utilize Ethereum’s existing infrastructure and tools with minimal modifications or adjustments.

To ensure the secure execution of transactions and smart contracts, Parallel Stack incorporates various security protocols and verification mechanisms, including transaction signature verification, smart contract auditing, and anomaly detection systems. To facilitate the development and deployment of applications on Parallel Stack, Sei Labs provides a comprehensive set of developer tools and APIs, aimed at helping developers fully leverage the high performance and scalability of Parallel Stack, thereby advancing the Ethereum ecosystem.

Innovating the Main Chain: a Polygon PoS Study in Parallelization

Author: Polygon Labs; Date: 2022.12.1

Polygon’s PoS chain has improved its transaction processing speed by 100% through the implementation of parallel EVM upgrades, thanks to the minimal metadata approach. Polygon adopted the principles of the Block-STM engine developed by Aptos Labs to create the minimal metadata method tailored to Polygon’s needs. The Block-STM engine is an innovative parallel execution mechanism that assumes no conflicts between transactions. During transaction execution, the Block-STM engine monitors each transaction’s memory operations, identifies and marks dependencies, and reorders conflicting transactions for validation to ensure result accuracy.

The minimal metadata method records dependencies of all transactions in the block and stores them in a Directed Acyclic Graph (DAG). Block proposers and validators first execute transactions, record dependencies, and attach them as metadata. When the block propagates to other nodes in the network, the dependency information is already included, reducing computational and I/O burdens for revalidation and enhancing verification efficiency. By recording dependencies, the minimal metadata method also optimizes transaction execution paths, minimizing conflicts.

What’s the Point of Parallelizing EVM? Or Is it the Endgame under EVM Dominance?

Author: Zhixiong Pan, founder of ChainFeeds; Date: 2024.3.28

Parallel EVM technology has garnered attention and investment from top venture capital firms, including Paradigm, Jump, and Dragonfly. These investors are optimistic about parallel EVM’s potential to break through the performance limitations of existing blockchain technologies, achieving more efficient transaction processing and broader application possibilities.

While the term “parallel EVM” literally signifies “parallelization,” it encompasses more than just enabling simultaneous processing of multiple transactions or tasks. It includes deep performance optimizations across various components of the Ethereum EVM, such as improving data access speed, increasing computational efficiency, and optimizing state management. Thus, these efforts likely represent the performance limits of the EVM standard.

In addition to technical challenges, parallel EVM faces issues in ecosystem building and market acceptance. It is essential to create differentiation within the open-source ecosystem and strike an appropriate balance between decentralization and high performance. Market acceptance requires demonstrating that the parallelization capabilities genuinely offer performance improvements and cost benefits, particularly in the context of existing Ethereum applications and smart contracts, which are already operating stably. Furthermore, promoting parallel EVM needs to address potential security issues and new technical flaws, ensuring system stability and user asset security—critical factors for widespread adoption of new technologies.

Death, Taxes, and EVM Parallelization

Author: Reforge Research; Date: 2024.4.1

The introduction of parallel EVM has improved the feasibility of on-chain Central Limit Order Books (CLOBs), with DeFi activity expected to increase significantly. In CLOBs, orders are sorted based on price and time priority, ensuring market fairness and transparency. However, implementing CLOBs on blockchain platforms like Ethereum often leads to high latency and transaction costs due to platform limitations in processing power and speed. The advent of parallel EVM has greatly enhanced the network’s processing capability and efficiency, enabling DeFi trading platforms to achieve faster and more efficient order matching and execution. Thus, CLOBs have become viable.

On this basis, Programmable Central Limit Order Books (pCLOBs) further extend CLOB functionality. pCLOBs not only provide basic buy and sell order matching features but also allow developers to embed custom smart contract logic during order submission and execution. This custom logic can be used for additional validation, execution condition determination, and dynamic adjustment of transaction fees. By embedding smart contracts in the order book, pCLOBs offer greater flexibility and security, supporting more complex trading strategies and financial products. Utilizing the high performance and parallel processing capabilities provided by parallel EVM, pCLOBs can achieve complex and efficient trading functions in a decentralized environment similar to traditional financial trading platforms.

However, despite significant improvements in blockchain performance due to parallel EVM, existing Ethereum Virtual Machine (EVM) and smart contract security still face shortcomings and are vulnerable to hacking. To address these issues, the author suggests adopting a dual VM architecture. In this architecture, in addition to the EVM, an independent virtual machine (e.g., CosmWasm) is introduced to monitor the execution of EVM smart contracts in real-time. This independent virtual machine functions similarly to antivirus software in an operating system, providing advanced detection and protection to reduce hacking risks. Emerging solutions like Arbitrum Stylus and Artela are considered promising for successfully implementing such a dual VM architecture. Through this architecture, these new systems can better embed real-time protection and other critical security features from the outset.

What Will Be the Next Step Towards Enhanced Scalability While Maintaining EVM Compatibility?

Author: Grace Deng, Researcher at SevenX Ventures; Date: 2024.4.5

New Layer 1 solutions like Solana and Sui offer higher performance than traditional Layer 2 and Layer 1 through the use of entirely new virtual machines (VMs) and programming languages, employing parallel execution, new consensus mechanisms, and database designs. However, these systems are not EVM-compatible, resulting in liquidity issues and higher barriers for users and developers. EVM-compatible Layer 1 blockchains like BNB and AVAX, despite improvements at the consensus layer, have made fewer modifications to the execution engine, leading to limited performance gains.

Parallel EVM can enhance performance without sacrificing EVM compatibility. For example, Sei V2 improves read and write efficiency by adopting optimistic concurrency control (OCC) and introducing a new state tree (IAVL trie); Canto Cyclone optimizes state management using the latest Cosmos SDK and ABCI 2.0 technologies, along with an in-memory IAVL state tree; and Monad proposes a new Layer 1 solution combining high throughput, decentralization, and EVM compatibility, utilizing OCC, new parallel-access databases, and a Hotstuff-based MonadBFT consensus mechanism.

Additionally, integrating other high-performance virtual machines (AltVMs) into the Ethereum ecosystem, particularly those supporting Rust development such as Solana’s Sealevel or Near’s WASM-based VM, could address the shortcomings of EVM incompatibility. This integration would not only overcome the issues but also attract Rust developers to the Ethereum ecosystem, enhancing overall performance and security while exploring new technological possibilities.

A Comprehensive Analysis of Parallel EVM: How to Overcome Blockchain Performance

Author: Gryphsis Academy; Date: 2024.4.5

Parallel EVM primarily focuses on optimizing the performance of the execution layer and is divided into Layer 1 and Layer 2 solutions. Layer 1 solutions introduce transaction parallel execution mechanisms, allowing transactions to be processed in parallel within the virtual machine. Layer 2 solutions essentially leverage the already parallelized Layer 1 virtual machines to achieve some degree of off-chain execution and on-chain settlement. In the future, the Layer 1 space may split into parallel EVM and non-EVM camps, while the Layer 2 space will evolve towards blockchain virtual machine simulators or modular blockchains.

Parallel execution mechanisms are mainly categorized into the following three types:

  1. Message Passing Model: Each actor can only access its own private data and must use message passing to access other data.
  2. Shared Memory Model: Uses memory locks to control access to shared resources, including memory lock models and optimistic parallelization.
  3. Strict State Access List: Based on the UTXO model, it precomputes the account addresses each transaction will access, forming an access list.

Different projects employ various strategies to implement parallel execution mechanisms:

  1. Sei v2: Transitions from a memory lock model to an optimistic parallelization model, optimizing potential data contention.
  2. Monad: Introduces superscalar pipelining technology and improved optimistic parallel mechanisms to achieve up to 10,000 TPS performance.
  3. Canto: Utilizes Cyclone EVM to introduce optimistic parallelization and innovates on decentralized financial infrastructure.
  4. Fuel: As a modular Ethereum rollup operating system, it adopts the UTXO model and optimistic parallelization mechanisms to increase transaction throughput.
  5. Neon, Eclipse, and Lumio: Provide cross-ecosystem performance improvements by integrating various Layer 1 chains, employing dual VM-supported strategies.

While parallel EVM offers an effective solution, it also introduces new security challenges. Parallel execution adds complexity due to multithreaded programming, leading to issues such as race conditions, deadlocks, livelocks, and starvation, which impact system stability and security. Additionally, new security vulnerabilities may arise, such as malicious transactions exploiting parallel execution mechanisms to create data inconsistencies or launch competitive attacks.

Disclaimer:

  1. This article is reprinted from [ChainFeeds Research]. All copyrights belong to the original author [0xNatalie]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
  2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
  3. Translations of the article into other languages are done by the Gate Learn team. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.

Parallel EVM Research Summary: Ecosystem and Features Explained

Intermediate8/27/2024, 1:20:53 AM
This article explores Parallel EVM technology, and its applications and potential in blockchain. It has attracted attention and investment from leading venture capital firms. Parallel EVM supports more complex and efficient blockchain applications through comprehensive performance optimization, achieving differentiation in the open-source ecosystem while balancing decentralization and high-performance needs. The article analyzes the technical challenges of Parallel EVM, such as data consistency, state access efficiency, and transaction conflict detection, and introduces unique solutions from projects like MegaETH and Artela.

TL;DR

  1. Parallel EVM has attracted the attention of top venture capital, and many projects have begun to explore this direction, such as Monad, MegaETH, Artela, BNB, Sei Labs, Polygon, etc.
  2. Parallel EVM is not just about achieving parallel processing but also involves comprehensive performance optimization of each component of the EVM. Through these optimizations, more complex and efficient blockchain applications are supported.
  3. Parallel EVMs must stand out in the open-source ecosystem by balancing decentralization and high performance, while also addressing potential security issues and market acceptance challenges. The complexity of multi-threaded programming introduces the challenge of managing multiple transactions at once, requiring effective solutions to ensure system stability and security.
  4. In the future, parallel EVM will promote the implementation of on-chain central limit order books (CLOB) and programmable central limit order books (pCLOB), which will greatly enhance the efficiency of DeFi activities, and the DeFi ecosystem is expected to grow significantly.
  5. Integrating other high-performance virtual machines (AltVM) into the Ethereum ecosystem will greatly enhance both performance and security. This approach leverages the strengths of each virtual machine, driving further development of Ethereum.

This year, Parallel EVM has garnered attention from leading venture capital firms like Paradigm and Dragonfly, capturing significant market interest. Unlike the traditional EVM, which processes transactions sequentially and can cause congestion and delays during peak periods, Parallel EVM leverages parallel processing technology to execute multiple transactions simultaneously, dramatically speeding up transaction processing. As complex applications such as on-chain games and account abstraction wallets become more prevalent, the demand for blockchain performance grows. To accommodate a larger user base, blockchain networks must handle high transaction volumes efficiently. Consequently, Parallel EVM is vital for the advancement of Web3 applications.

However, implementing Parallel EVM comes with common challenges that require precise technical solutions to ensure stable system operation.

  • Data Consistency: In Parallel EVM, multiple transactions occurring simultaneously may require reading or modifying account information concurrently. Effective locking mechanisms or transaction processing methods are needed to ensure data consistency during state modifications.
  • State Access Efficiency: Parallel EVM must quickly access and update states, demanding efficient state storage and retrieval mechanisms. Optimizing storage structures and access paths, such as using advanced data indexing techniques and caching strategies, can significantly reduce data access latency and enhance overall system performance.
  • Transaction Conflict Detection: In parallel execution, multiple transactions may depend on the same data state, making transaction ordering and dependency management complex. Complex scheduling algorithms are required to identify and manage dependencies between parallel transactions, detect potential conflicts, and decide on handling methods to ensure parallel execution results are consistent with serial execution.

For example, MegaETH decouples transaction execution tasks from full nodes, assigning different tasks to specialized nodes to optimize overall system performance. Artela uses predictive optimistic execution and asynchronous preloading technologies to analyze transaction dependencies with AI and preload required transaction states into memory, improving state access efficiency. BNB Chain developed specialized conflict detectors and re-execution mechanisms to enhance transaction dependency management, reducing unnecessary re-executions, etc.

To understand the development direction of Parallel EVM in depth, here are nine selected high-quality articles on the topic, providing comprehensive perspectives on different chains’ implementation plans, ecosystem studies, and future prospects.

MegaETH: Unveiling the First Real-Time Blockchain

Author: MegaETH; Date: June 27, 2024

MegaETH is an EVM-compatible Layer 2 aiming to achieve near-Web2 server real-time performance. Its goal is to push Ethereum L2 performance to hardware limits, offering high transaction throughput, ample computing power, and millisecond response times. This allows developers to build and combine complex applications without performance constraints.

MegaETH enhances performance by separating transaction execution tasks from full nodes and introducing parallel processing technology. Its architecture consists of three main roles: Sequencer, Validator, and Full Node.

  • Sequencer: Responsible for ordering and executing user-submitted transactions. After executing transactions, Sequencers send state changes (state diffs) to Full Nodes via a peer-to-peer (p2p) network.
  • Full Node: Receives state diffs from Sequencers and directly applies these changes to update its local blockchain state, avoiding transaction re-execution. This significantly reduces computational resource consumption and improves overall system efficiency.
  • Validator: Uses stateless verification schemes to validate blocks, allowing simultaneous verification of multiple blocks. This further enhances verification efficiency and speed.

This specialized node design allows different types of nodes to set independent hardware requirements based on their functions. For instance, Sequencers need high-performance servers to handle a large volume of transactions, while Full Nodes and Validators can use relatively lower-spec hardware.

Presenting Artela Scalability Whitepaper - Parallel Execution Stack and Elastic Block Space

Author: Artela; Date: 2024.6.20

Artela significantly enhances blockchain parallel execution efficiency and overall performance through several key technologies:

  1. Parallel Execution: By predicting transaction dependencies and grouping transactions, it uses multiple CPU cores to parallel process, improving computational efficiency.
  2. Parallel Storage: Optimizes the storage layer to support parallel data processing, avoiding storage bottlenecks and enhancing overall system performance.
  3. Elastic Computing: Supports multiple computers working together, creating elastic computing nodes and block space, providing higher transaction throughput and predictable performance for dApps.

Specifically, Artela’s predictive optimistic execution uses AI to analyze dependencies between transactions and contracts, predicting potential conflicting transactions and grouping them to reduce conflicts and re-executions. The system dynamically accumulates and stores historical transaction state access information for predictive algorithms. Asynchronous preloading loads required transaction states into memory to avoid I/O bottlenecks during execution. Parallel storage improves Merkleization and I/O performance by separating state commitments from storage operations, managing parallel and non-parallel operations independently to further enhance parallel efficiency.

Additionally, Artela’s elastic computing builds elastic block space (EBS). Traditional blockchains share a single block space among all dApps, leading to resource competition among high-traffic dApps, causing unstable gas fees and unpredictable performance. Elastic block space provides dedicated and dynamically scalable block space for dApps, ensuring predictable performance. dApps can apply for exclusive block space as needed, and as block space increases, Validators can extend processing capabilities by adding elastic execution nodes, ensuring efficient resource utilization and adapting to different transaction volumes.

Road to High Performance: Parallel EVM for BNB Chain

Author: BNB Chain; Date: 2024.2.16

On the BNB Chain, the team has taken several steps to achieve Parallel EVM to enhance transaction processing capacity and scalability. Key developments include:

Parallel EVM v1.0:

  • Scheduler: Allocates transactions to different threads for parallel execution to optimize throughput.
  • Parallel Execution Engine: Utilizes parallel processing on dedicated threads to execute transactions independently, significantly reducing processing time.
  • Local State Database: Each thread maintains its own “thread-local” state database to efficiently record state access information during execution.
  • Conflict Detection and Re-execution: Ensures data integrity by detecting and managing transaction dependencies, re-executing transactions in case of conflicts to ensure accuracy.
  • State Submission Mechanism: Once transactions are executed, results are seamlessly submitted to the global state database to update the overall blockchain state.

Parallel EVM v2.0

Based on parallel EVM 1.0, the BNB chain community has introduced a series of innovations to improve performance:

  • Streaming Pipeline: Enhances execution efficiency, allowing smooth transaction processing in the parallel engine.
  • Universal Unconfirmed State Access: Optimizes state information access by allowing other transactions to temporarily use results from unconfirmed transactions, reducing inter-transaction wait times.
  • Conflict Detector 2.0: Improved conflict detection mechanism for better performance and accuracy, ensuring data integrity while reducing unnecessary re-executions.
  • Scheduler Enhancements: Utilizes static and dynamic scheduling strategies for more efficient workload allocation and resource optimization.
  • Memory Optimization: Significantly reduces memory usage through shared memory pools and lightweight copy techniques, further improving system performance.

Parallel EVM v3.0

After the performance improvements of parallel EVM 2.0, the BNB chain community actively developed parallel EVM 3.0 with the following goals:

  • Reduce or Eliminate Re-execution: Introduces a Hint-based scheduler using external hint providers to analyze transactions and predict potential state access conflicts. This helps better schedule transactions and reduce conflicts, minimizing re-execution needs.
  • Modularity: Decomposes code into independent modules for better maintainability and adaptation to different environments.
  • Codebase Refactoring: Aligns with the latest BSC/opBNB codebase to ensure compatibility and simplify integration.
  • Thorough Testing and Validation: Conducts extensive testing under various scenarios and workloads to ensure solution stability and reliability.

Sei’s Parallel Stack

Author: Sei; Date: 2024.3.13

Sei Labs has created an open-source framework called Parallel Stack, designed to build Layer 2 solutions that support parallel processing technology. The core advantage of Parallel Stack lies in its parallel processing capability, leveraging advancements in modern hardware to reduce transaction costs. This framework employs a modular design, allowing developers to add or modify functionality modules based on specific needs, thereby adapting to various application scenarios and performance requirements. Parallel Stack can seamlessly integrate with the existing Ethereum ecosystem, enabling applications and developers to utilize Ethereum’s existing infrastructure and tools with minimal modifications or adjustments.

To ensure the secure execution of transactions and smart contracts, Parallel Stack incorporates various security protocols and verification mechanisms, including transaction signature verification, smart contract auditing, and anomaly detection systems. To facilitate the development and deployment of applications on Parallel Stack, Sei Labs provides a comprehensive set of developer tools and APIs, aimed at helping developers fully leverage the high performance and scalability of Parallel Stack, thereby advancing the Ethereum ecosystem.

Innovating the Main Chain: a Polygon PoS Study in Parallelization

Author: Polygon Labs; Date: 2022.12.1

Polygon’s PoS chain has improved its transaction processing speed by 100% through the implementation of parallel EVM upgrades, thanks to the minimal metadata approach. Polygon adopted the principles of the Block-STM engine developed by Aptos Labs to create the minimal metadata method tailored to Polygon’s needs. The Block-STM engine is an innovative parallel execution mechanism that assumes no conflicts between transactions. During transaction execution, the Block-STM engine monitors each transaction’s memory operations, identifies and marks dependencies, and reorders conflicting transactions for validation to ensure result accuracy.

The minimal metadata method records dependencies of all transactions in the block and stores them in a Directed Acyclic Graph (DAG). Block proposers and validators first execute transactions, record dependencies, and attach them as metadata. When the block propagates to other nodes in the network, the dependency information is already included, reducing computational and I/O burdens for revalidation and enhancing verification efficiency. By recording dependencies, the minimal metadata method also optimizes transaction execution paths, minimizing conflicts.

What’s the Point of Parallelizing EVM? Or Is it the Endgame under EVM Dominance?

Author: Zhixiong Pan, founder of ChainFeeds; Date: 2024.3.28

Parallel EVM technology has garnered attention and investment from top venture capital firms, including Paradigm, Jump, and Dragonfly. These investors are optimistic about parallel EVM’s potential to break through the performance limitations of existing blockchain technologies, achieving more efficient transaction processing and broader application possibilities.

While the term “parallel EVM” literally signifies “parallelization,” it encompasses more than just enabling simultaneous processing of multiple transactions or tasks. It includes deep performance optimizations across various components of the Ethereum EVM, such as improving data access speed, increasing computational efficiency, and optimizing state management. Thus, these efforts likely represent the performance limits of the EVM standard.

In addition to technical challenges, parallel EVM faces issues in ecosystem building and market acceptance. It is essential to create differentiation within the open-source ecosystem and strike an appropriate balance between decentralization and high performance. Market acceptance requires demonstrating that the parallelization capabilities genuinely offer performance improvements and cost benefits, particularly in the context of existing Ethereum applications and smart contracts, which are already operating stably. Furthermore, promoting parallel EVM needs to address potential security issues and new technical flaws, ensuring system stability and user asset security—critical factors for widespread adoption of new technologies.

Death, Taxes, and EVM Parallelization

Author: Reforge Research; Date: 2024.4.1

The introduction of parallel EVM has improved the feasibility of on-chain Central Limit Order Books (CLOBs), with DeFi activity expected to increase significantly. In CLOBs, orders are sorted based on price and time priority, ensuring market fairness and transparency. However, implementing CLOBs on blockchain platforms like Ethereum often leads to high latency and transaction costs due to platform limitations in processing power and speed. The advent of parallel EVM has greatly enhanced the network’s processing capability and efficiency, enabling DeFi trading platforms to achieve faster and more efficient order matching and execution. Thus, CLOBs have become viable.

On this basis, Programmable Central Limit Order Books (pCLOBs) further extend CLOB functionality. pCLOBs not only provide basic buy and sell order matching features but also allow developers to embed custom smart contract logic during order submission and execution. This custom logic can be used for additional validation, execution condition determination, and dynamic adjustment of transaction fees. By embedding smart contracts in the order book, pCLOBs offer greater flexibility and security, supporting more complex trading strategies and financial products. Utilizing the high performance and parallel processing capabilities provided by parallel EVM, pCLOBs can achieve complex and efficient trading functions in a decentralized environment similar to traditional financial trading platforms.

However, despite significant improvements in blockchain performance due to parallel EVM, existing Ethereum Virtual Machine (EVM) and smart contract security still face shortcomings and are vulnerable to hacking. To address these issues, the author suggests adopting a dual VM architecture. In this architecture, in addition to the EVM, an independent virtual machine (e.g., CosmWasm) is introduced to monitor the execution of EVM smart contracts in real-time. This independent virtual machine functions similarly to antivirus software in an operating system, providing advanced detection and protection to reduce hacking risks. Emerging solutions like Arbitrum Stylus and Artela are considered promising for successfully implementing such a dual VM architecture. Through this architecture, these new systems can better embed real-time protection and other critical security features from the outset.

What Will Be the Next Step Towards Enhanced Scalability While Maintaining EVM Compatibility?

Author: Grace Deng, Researcher at SevenX Ventures; Date: 2024.4.5

New Layer 1 solutions like Solana and Sui offer higher performance than traditional Layer 2 and Layer 1 through the use of entirely new virtual machines (VMs) and programming languages, employing parallel execution, new consensus mechanisms, and database designs. However, these systems are not EVM-compatible, resulting in liquidity issues and higher barriers for users and developers. EVM-compatible Layer 1 blockchains like BNB and AVAX, despite improvements at the consensus layer, have made fewer modifications to the execution engine, leading to limited performance gains.

Parallel EVM can enhance performance without sacrificing EVM compatibility. For example, Sei V2 improves read and write efficiency by adopting optimistic concurrency control (OCC) and introducing a new state tree (IAVL trie); Canto Cyclone optimizes state management using the latest Cosmos SDK and ABCI 2.0 technologies, along with an in-memory IAVL state tree; and Monad proposes a new Layer 1 solution combining high throughput, decentralization, and EVM compatibility, utilizing OCC, new parallel-access databases, and a Hotstuff-based MonadBFT consensus mechanism.

Additionally, integrating other high-performance virtual machines (AltVMs) into the Ethereum ecosystem, particularly those supporting Rust development such as Solana’s Sealevel or Near’s WASM-based VM, could address the shortcomings of EVM incompatibility. This integration would not only overcome the issues but also attract Rust developers to the Ethereum ecosystem, enhancing overall performance and security while exploring new technological possibilities.

A Comprehensive Analysis of Parallel EVM: How to Overcome Blockchain Performance

Author: Gryphsis Academy; Date: 2024.4.5

Parallel EVM primarily focuses on optimizing the performance of the execution layer and is divided into Layer 1 and Layer 2 solutions. Layer 1 solutions introduce transaction parallel execution mechanisms, allowing transactions to be processed in parallel within the virtual machine. Layer 2 solutions essentially leverage the already parallelized Layer 1 virtual machines to achieve some degree of off-chain execution and on-chain settlement. In the future, the Layer 1 space may split into parallel EVM and non-EVM camps, while the Layer 2 space will evolve towards blockchain virtual machine simulators or modular blockchains.

Parallel execution mechanisms are mainly categorized into the following three types:

  1. Message Passing Model: Each actor can only access its own private data and must use message passing to access other data.
  2. Shared Memory Model: Uses memory locks to control access to shared resources, including memory lock models and optimistic parallelization.
  3. Strict State Access List: Based on the UTXO model, it precomputes the account addresses each transaction will access, forming an access list.

Different projects employ various strategies to implement parallel execution mechanisms:

  1. Sei v2: Transitions from a memory lock model to an optimistic parallelization model, optimizing potential data contention.
  2. Monad: Introduces superscalar pipelining technology and improved optimistic parallel mechanisms to achieve up to 10,000 TPS performance.
  3. Canto: Utilizes Cyclone EVM to introduce optimistic parallelization and innovates on decentralized financial infrastructure.
  4. Fuel: As a modular Ethereum rollup operating system, it adopts the UTXO model and optimistic parallelization mechanisms to increase transaction throughput.
  5. Neon, Eclipse, and Lumio: Provide cross-ecosystem performance improvements by integrating various Layer 1 chains, employing dual VM-supported strategies.

While parallel EVM offers an effective solution, it also introduces new security challenges. Parallel execution adds complexity due to multithreaded programming, leading to issues such as race conditions, deadlocks, livelocks, and starvation, which impact system stability and security. Additionally, new security vulnerabilities may arise, such as malicious transactions exploiting parallel execution mechanisms to create data inconsistencies or launch competitive attacks.

Disclaimer:

  1. This article is reprinted from [ChainFeeds Research]. All copyrights belong to the original author [0xNatalie]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
  2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
  3. Translations of the article into other languages are done by the Gate Learn team. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.
Start Now
Sign up and get a
$100
Voucher!