Aptos is a highly scalable and decidedly modular Layer 1 blockchain that leverages many novel approaches to transaction flow, parallel execution, and the transaction processing lifecycle to achieve high levels of throughput and low latency.

Summary

Aptos is built in Move and designed to be resolutely modular and seamlessly upgradeable. In addition, the protocol has created several features to improve user experience while simultaneously improving security, such as pre-signing transaction transparency that describes the literal outcome of a transaction in plain language before a user signs it as well as hybrid key management systems that boost security and helps reduce key mismanagement risks.

Overview

Mainstream Web3 adoption is under way. Activity across blockchains has grown considerably over the last several years and new decentralized applications (dApps) along with their accompanying use cases continue to emerge. Core infrastructure that supports these networks and functions without requiring trust in a centralized entity is becoming increasingly robust. Yet solutions that are ultra performant and truly enterprise-grade ready are few and far between. In general, part of the issue stems from blockchain’s current lack real-time responsiveness and throughput limitations. This is further compounded by the high costs, security risks, inconsistent performance, and scaling issues that persist in most Layer 1 chains. The end result is that real world use cases become inherently constrained.

Many emerging blockchains vie to confront these challenges, few have delivered. Aptos, however, takes a different path and leverages novel approaches to transaction flow, computation, and smart contract development. The end result is a blockchain that is engineered with the potential to optimize scalability while reinforcing security. Aptos also has a decidedly modular architecture which enables it to be flexible and easily upgradable so that users can quickly adapt to changes throughout Web3 with less friction and downtime. Collectively, these features can help Aptos deploy a network which supports mass adoption and overcomes many of the hurdles that have hindered the development of other Layer 1 networks.

Apotos confronts these issues from a truly bottoms-up approach, starting with the very programming language itself. The protocol is built on Move, a highly flexible language inspired by Rust, which has the potential to natively optimize transaction speeds and achieve higher security. Move also enables valuable user features, including the Move Prover, an automatic formal verification system that detects potentially malicious smart contracts and acts as an inherent safeguard in smart contract writing. In addition, Move’s data model facilitates more flexible private key management and hybrid custody options, along with transaction transparency prior to signing. So Move comes with a variety of built-in features and additional layers of security which contribute to a safer, more seamless user experience.

Beyond the programming language, transaction flow on Aptos is resolutely different than most competing networks. Each stage of the process, such as broadcasting transactions, ordering block metadata, and batching storage, is completely modular and operates concurrently. One reason this design is important is that it optimizes all available computing resources and unlocks a higher level of parallel execution. Furthermore, Aptos’ parallel execution engine does not require a priori knowledge of data locations to be read and written. This removes a key limitation that exists on most protocols that utilize parallel execution and enables faster throughput and less latency. Finally, Aptos achieves horizontal scaling by enabling internal sharding of validators within the network as well as homogenous state sharding. Collectively, these designs provide Aptos with the potential to drive a truly enterprise-grade blockchain network.

Screen Shot 2022 08 12 at 2 19 33 PM

Deeper Dive: Consensus Mechanism, Network Validators, and State Synchronization

The Aptos blockchain leverages a proof-of-stake (PoS) consensus mechanism and is powered by a network of validators which process transactions and update the system. Similar to other PoS blockchains, the consensus voting power of each validator within the network is proportional to the amount of Aptos tokens staked within them. Validators that act maliciously or suffer downtime can be slashed, so token holders are incentivized to stake their tokens in validators that are honest and performant. The protocol implements a system of rapid, stake-weight rotation to optimize validator performance as well as facilitate organic decentralization of the network, as validators are constantly rotating, rather than operating as a static or slow moving set. In addition, and on-chain reputation tracking mechanism is used promote performant validators (e.g. validators that have successfully submitted a block within a given time) and demote non-performant validators (e.g. validators that have gone offline). This mechanism helps improve incentive structures for validators and minimizes the impact of poorly performing validators.

Nodes in the network, which submit transactions and query blockchain history, can exist in either a full or light version. Full nodes are responsible for a much heavier workload as they verify the entire transaction and state history from validators or other full nodes in the network. One interesting aspect of full nodes on Aptos is that they have the discretion to remove data history from their memory. This is important for many reasons, but particularly in relation to performance and scaling, as the ability to truncate data assists in minimizing storage demands. Light clients, on the other hand, are only required to maintain the current set of validators in order to query partial blockchain history data, making their workload significantly less rigorous.

To help maximize node flexibility, Aptos implements a unique approach to state synchronization. In short, a state synchronization protocol contains the rules which govern blockchain data across the network and ensure that all participants in the system are synchronized. Aptos allows participants to select different strategies that best fit their intended use case and which account for differences in resource availability. For instance, a full node can choose to synchronize the full history of the blockchain, leading back to the very first transaction, or it can choose to synchronize the most recent state of the blockchain, while light clients can simply choose to synchronize with partial state histories. The key takeaway is that Aptos leverages a state synchronization protocol that is highly customizable and lets each participant choose which part of the continuum of blockchain data that they want to process and retain. Developing a protocol that is flexible and efficient is critical for adopting to a diverse set of needs and use cases, as well as for achieving high throughput for each participant in the system. Overall, light clients, full nodes, and validators all can benefit from more efficient ways of sending, receiving, and verifying blockchain data, which in turn leads to better user experiences.

Design Principles and the Move Language

Shifting back to the idea of programming language and protocol design, Move has an interesting appeal in that its core focus is around safety and flexibility, from both a universal protocol level and a user experience perspective. This extends in many meaningful ways by which Move is integrated into Aptos, including to perform transaction execution and to represent blockchain state through its object model. Moreover, Move emphasizes access control and resource scarcity. One reason this is important is that it includes native safeguards against attack vectors like double-spending. The language also has a set of powerful developer tools such as the Move virtual machine and a formal verifier, Move Prover. The Move Prover confirms the correctness of code being written by allowing it to be tested against a given condition and ensures that the resulting output is indeed the expected output. This creates a more secure environment as well as a variety of quality of life improvements for developers and end users alike.

Move modules encode the rules governing the state transitions of the blockchain. Users can submit transactions that can update existing code, publish new code, or interact with existing modules. So there is tremendous power and function built directly within the language itself. Overall, Move modules define the rules and every element within the system, such as storage and access, which is fundamentally important in that it ensures that critical resources like tokens cannot be double-spent.

Another interesting feature of Move is that enables user accounts on Aptos are protected by a variety privacy-preserving primitives. Users can create multiple accounts without any inherent link to each other, all while using a single wallet to control each of these different accounts. This creates a user-friendly asset management environment while introducing improved privacy and censorship-resistance.

The protocol also includes a variety of improved user experience features. Transaction viability protection mitigates the risk of a user unintentionally signing a transaction by constraining the viability of every transaction and protecting the user from unbounded validity. The Move-based key management allows users to delegate their private key to one or more custodians and other trusted entities and rotate the key according to specific circumstances. Pre-signing transaction transparency helps reduce the risk of a user signing malicious transactions by providing literal description of the outcomes of their transactions prior to signing. Attack history data and malicious smart contracts can a be fed into this function to further reduce fraud risk.

Ledger State and Data Model

The ledger state represents the state of every account on the Aptos blockchain and is updated continuously, as the version of the state at any given time is correlated to the number of transactions the system has executed at that time. One key feature of the ledger state composition is the way that it handles transaction events. In short, each transactions is stored within the ledger and has a corresponding unique key to allow nodes to query it. Transactions can only generate events and cannot read events. This means that transaction execution is a function of the current state, and not the historic state.

Each account on the ledger has a unique account address that are generated by a signature key-pair. An important features of this approach is that Aptos allows users to create multiple accounts by generating multiple key-pairs. These new accounts have no inherent link to each other, yet the user can manage them all in a single wallet. This creates privacy-protecting primitives and a cleaner user experience where users can seamlessly manage numerous accounts through one wallet interface.

The Transaction Flow and Lifecycle

Aptos employs a novel approach to the transaction flow by delineating pipelining, batching, and parallel execution into their own separate stages, each completely independent of the other. This helps maximize throughput and resource utilization, while simultaneously reducing complexity. In addition, this design can improve development cycles insofar as each stage can be targeted as its own independent entity rather than as one indivisible, inner-connect structure.

Screen Shot 2022 08 12 at 4 27 59 PM

One of the key differentiating features is that the consensus phase is decoupled from non-agreement types of tasks. This means that tasks such as transaction dissemination and transaction execution function outside of the consensus phase, which reduces operational complexity and allows block metadata and proof ordering to function correctly with much less bandwidth demands than existing schemas. Ultimately, this contributes to the network’s ability to increase throughput and minimize latency.

Improved User Experience

Aptos implements several features to improve user experience. First, the protocol places inherent constraints on the viability of every transaction on the network in way that mitigates the risk of a user transaction being manipulated, such as a replay attack. Second, Move enables a flexible key management system that creates hybrid models of custody, such as rotating private keys to a set of trusted custodial entities along with on-chain key recovery mechanisms. Third, Aptos implements a system of transaction transparency, in which users can see exactly what their transaction will yield before it is executed. This includes the literal outcomes of the transaction, in easy to read language, as well as built-in wallet parameters that automatically reject specific transaction conditions, such as sending to a contract that has been identified as malicious.

Parallel Execution

A core tenet of Aptos’ design is to maximize parallel execution, wherever possible. Thus, the protocol implements a parallel data model as well as a parallel execution engine. In terms of the data model, any transaction that exists without a conflict in data or account can be executed in parallel. In terms of the execution engine, conflicts are pro-actively managed and transactions are processed, optimistically, in batches, and then verified after execution. Compared with sequential execution, Aptos’ engine can perform up to 16 times faster. In comparison with other parallel execution engines, Aptos’ iteration allows for more complex transactions to occur which helps decrease the costs and latency for end users using the network.

Mechanism Design and Tokenomics

One of the more interesting applications of the native token is around its role in facilitating optimal transaction flow. All transactions on the network require a specific amount of gas in order to complete, paid in $APTOS, and validators in the network can utilize the gas fee parameter as a proxy for transaction prioritization. That is, validators can easily filter transactions according to their value. This filtering can be done at any stage of the transaction processing lifecycle. Perhaps just as importantly, however, is that low-value transactions can be identified and discarded at any stage of the processing lifecycle as well. This allows the network to operate efficiently during times of heavy traffic. So again, transaction flow on Aptos is composed in a very modular way, rather than as a monolith.

Macro Perspective

Aptos joins a consortium of competing Layer 1 networks that are increasingly becoming more competitive and performant. One its key differentiating features is its development in Move. The reason this is important is that Move gives developers the potential to design smart contracts in novel ways, many of which result in significant gains in performance and security as well as user experience. The ability to manage several private keys from a single user account, for instance, and utilize built-in transaction protection mechanisms is incredibly valuable for mass adoption. One general concern, however, is that while Move has been developed over the course of the last few years, it has yet to be deployed and battle-tested in a mainnet environment.

Interoperability, core infrastructure, and developer buy-in will be especially critical to the success of the protocol and ultimately, to how well it can integrate into the broader Web3 ecosystem. To facilitate interoperability, instances of bridges, such as Wormhole, are being deployed on Aptos. This will enable native users to move between different networks and also allow other users from other blockchains such as Ethereum and Solana to enter the Aptos ecosystem more seamlessly. In terms of infrastructure, Aptos is leveraging data oracles including Pyth to help service its native dApps and connect them with off-chain data.

Ultimately, the success or failure of any blockchain will rely on user adoption. No matter how fast transactions can be completed, how efficiently the network can scale, or how cheap using the network is, if users do not use it, then it will not succeed. Part of the way user adoption is impacted is through the development of dApps built on the network that have real world use cases. So a critical aspect of any blockchain is getting meaningful dApps built. This requires buy-in from talented developers. Aptos sets the stage with a flexible protocol design and an equally flexible programming language. If developers can be successfully onboard to the network, they greeted with potentially massive gains in execution, security, and usability. These features will play a critical role in Aptos’ positioning within the Layer 1 space.

Further Reading and Resources

  1. https://github.com/aptos-labs/aptos-core/blob/main/developer-docs-site/static/papers/whitepaper.pdf
  2. https://arxiv.org/pdf/2110.08362.pdf
  3. https://github.com/solana-labs/move
  4. https://aptoslabs.medium.com/

MetaMask Institutional offers unrivaled access to the DeFi ecosystem without compromising on institution-required security, operational efficiency, or compliance requirements. We enable funds to trade, stake, borrow, lend, invest, and interact with over 17,000 DeFi protocols and applications.


Found this research useful? Connect with the ConsenSys Cryptoeconomic Research team at [email protected]

Disclaimer: ConsenSys Software Inc. is not a registered or licensed advisor or broker. This report is for general informational purposes only.  It does not constitute or contain any individual investment advice and is made without any regard to the recipient’s objectives, financial situation, or means.  It is not an offer to buy or sell, or a solicitation of any offer to buy, any token or other investment, nor is it intended to be used for marketing purposes to anyone in any jurisdiction. ConsenSys does not intend for any person or entity to rely on any facts, opinions, or ideas, and any financial or economic commentary expressed in this report may not be relied upon. ConsenSys makes no representations as to the accuracy, completeness, or timeliness of the information or opinions in this report and, along with its employees, does not assume any responsibility for any loss to any person or entity that may result from any act or omission based upon this report. This report is subject to correction, completion, and amendment without notice; however, ConsenSys has no obligation to do so.

What's your reaction?

Leave a comment

Minimum 4 characters