Aptos — the new kid on the block
There is a new kid on the block and its’ after crypto mass adoption: Aptos.
In this post, we’d like to shed some light on the Aptos network, some of its’ USPs, share the reasons behind our supporting it as well as address a few FAQs on all things staking on the network.
Introduction — a brief trip down memory lane
Aptos is a fairly new layer-one blockchain initiated by Aptos Labs, a venture founded by former Diem engineers and scientists to develop ‘a Layer 1 for everyone’. Diem as you might remember, is Metas’ (formerly Facebook) hampered attempt to build their own, permissioned blockchain.
While Diem (formerly Libra) was quite a controversial and polarizing endeavor, it brought to light a new, Rust-inspired, low-level, open-source programming language called Move. The team behind Aptos was heavily involved in the development of Move, which was initiated with a focus on flexibility, customizable transaction logic, and leveraging provability to increase safety while writing smart contracts. It is therefore no surprise that Aptos is built with Move, but we’re getting ahead of ourselves. Let’s start from the top.
What is Aptos?
A textbook definition of the Aptos blockchain would be:
Aptos is a Byzantine-Fault-Tolerant (BFT), Proof-of-Stake (PoS), monolithic layer-one blockchain built for upgradability, safety, and performance.
A lot to process, we know — let’s unpack this one by one:
> Byzantine-Fault-Tolerant (BFT) and Proof-of-Stake (PoS)
To reach consensus and hence agree on the order and finalization of transactions, Aptos uses a BFT algorithm, named AptosBFT. Without going into detail here, BFT algorithms are found in many popular blockchains such as Solana or Cosmos. BFT describes a distributed computer’s network’s ability to operate despite malfunctioning or even malicious nodes — so preventing individual faults from contaminating the entire network. Here’s a great explainer video by Binance Academy that goes a bit more into detail.
Like most popular blockchains, Aptos also utilizes Proof-of-Stake to incentivize honest node participation and manage the maintenance of its’ ledger — you can learn more on PoS here. Together, BFT and PoS power blockchain networks to function without the presence of a centralized watchdog.
> Monolithic layer-one
Roughly speaking, you can divide (most) blockchains into two categories: modular blockchains vs. monolithic blockchains. However, note that this is to be seen as a spectrum rather than something binary.
Examples of the former are Polkadot and Ethereum. In these networks, ‘sharding’ plays a big role, which is the process of splitting up the blockchain network into multiple blockchains called shards, rollups, or sidechains (there are many terms out there). It is a way of partitioning the computational and storage workload to increase the scale of the overall network. Consensus, data availability, execution, and settlement are core components of blockchain architecture, which are spread out to at least two different layers in modular blockchain networks. While scale and throughput increase, composability might suffer as for certain types of use cases, different shards have to communicate. This might lead to compromises i.t.o. UX.
Monolithic blockchains are still rare and Solana is definitely the poster child for these kinds of chains. Rather than separating consensus, data availability, execution, and settlement, monolithic blockchains do it all in one layer, the layer-one or base-layer. This obviously introduces challenges i.t.o. congestion, network outages, or scalability due to increased load, which chains such as Solana or Aptos address with parallel execution (more on that later). However, it does heavily favor composability and hence, UX.
On the spectrum between modular and monolithic chains, Aptos is heavily tilted towards monolithic, however, it also has elements that remind one of a modular chain. They are mixing different types of sharding. One such is vertical sharding, i.e. validators running more powerful clusters of highly performant hardware. They also describe their approach as a ‘homogenous sharding platform’ in which users and developers can decide on how their data is sharded while having one single infrastructure, token, and language.
We are still early, so it is really hard to tell which approach is ‘better’ and will prevail or if there is justification for both approaches to co-exist in harmony (we are proponents of the latter). Aptos is a great example of this as it mixes approaches from both worlds. Here’s a great article that goes a bit deeper into the distinction between modular vs. monolithic chains.
Technical caveats & USPs of Aptos
The second half of our Aptos definition is also a good starting point for some of Aptos’ USPs and technical caveats. Let’s dive in:
Evolution taught us one thing, we need to adapt. This also holds true for blockchains, a) given their infancy and b) their objective of becoming the backend of web3, eventually handling trillions of value and billions of transactions.
For Aptos to constantly evolve and adapt, so it heavily emphasizes upgradeability to be able to do so. Upgradeability on Aptos is achieved through a multifaceted approach.
First, on a technological level: the configurations of the blockchain are stored on-chain, so changes to it are spread across the entire network with every state synchronization. For node operators like us, Aptos allows choosing a state synchronization strategy that best fits their needs, use case, and resource allocation preference. Aptos’ state synchronization protocol (the rules that govern how data is spread across the network & ensure that everyone is in sync re. the status quo of the network) is customizable in so far as it lets participants choose which blockchain data they want to process and retain. This allows participants (validators, full nodes, light clients) to choose how they want to send, receive, and verify blockchain data so that it can be best tailored to different use cases and, hence, further increase the network’s end user and developer experience. In addition, this feeds into seamless upgradeability for the network.
Furthermore, Aptos will have on-chain governance. That empowers $APT holders to influence the evolution of the network by submitting and voting on proposals. Furthermore, it aligns incentives and minimizes the cost of blockchain upgrades.
Before a proposal is moved on-chain, it is discussed in the Aptos forum as an AIP (Aptos Improvement Proposal). Proposers and voters need to stake their token and there is a minimum required proposer stake for anyone looking to make proposals as well as a lockup period. Both requirements are in place to prevent spam and guarantee a certain quality of proposals.
“Aptos is almost entirely configurable via an on-chain framework for making changes to the network” (Source). Hence, many parameters are subject to governance, such as the minimum & maximum amount of stake for validator nodes (more below), epoch duration, bug fixes, core code adjustments, as well as adding new functionality. The second half of our Aptos definition is also a good starting point for some of Aptos’ USPs and technical caveats. Let’s dive in:
To reach mass adoption, user experience (UX) plays an important role. A great UX heavily relies on two factors: security & performance. A roadblock for the former was key management for novel crypto users — although admittedly, also experienced users oftentimes long for better solutions.
Aptos wants to improve UX through flexible private key management and hybrid custody services. Powered through Move, Aptos accounts support key rotation. This means that users can create different accounts that are not inherently linked to one another, while at the same time using a single wallet to manage them. Furthermore, it enables hybrid key custody services, which allow users to appoint one or more trusted custodians, that can rotate the key private according to certain circumstances (e.g. N-of-M multi-sig) in case something goes south. This mutes the damage through key mismanagement and having it implemented on-chain at the base layer circumvents off-chain, differing solutions by wallet providers.
The ‘pre-signing transaction transparency’ describes the outcome of a transaction in human-readable form before a user signs it. Paired with specific transaction conditions that can automatically be rejected if these were flagged prior as malicious, it further increases end-user security and ultimately improve UX by an order of magnitude.
For developers and entrepreneurs, it is paramount to be able to write secure smart contracts so that they can provide their users with a great UX as well. Since Move was inspired by Rust, it introduced a new type of digital object, that is no longer stored in hashmaps, but can rather be passed in arguments and returned by functions, smart contract developer interactions are streamlined. This prevents dangling references and memory leaks while at the same time providing referential transparency for immutable references. Additionally, the Move Prover, a formal verifier, enables developers to verify that the code actually does what it was intended to do before anything has to be deployed to a live environment.
The show pony for Aptos i.t.o. performance is a concept called ‘parallel execution’ — brought to light and popularized first and foremost by Solanas’ sealevel runtime. There are two approaches to handling transactions, sequentially (i.e. one after another) or in parallel (you guessed it).
The parallel processing of transactions in Aptos is done through the Block-STM execution engine. Broken down, this engine facilitates that any transaction can be processed in parallel as long as it does not exist with a conflict in data or account. It operates and processes transactions optimistically in batches, meaning that batches of transactions are validated POST-execution and re-executed in case they were invalid. Block-STM allows devs to build products that process multiple, non-overlapping transactions at once, allowing scale on a monolithic chain.
The transaction lifecycle in Aptos is modular: broadcasting, ordering block metadata, and batching storage each operating separately and concurrently. This empowers the system to best leverage all available computing resources and, hence, reinforce the parallel execution of transactions.
Furthermore, the consensus phase takes place separately from transaction dissemination and execution, so that these can happen outside the consensus phase, ultimately decreasing bandwidth requirements, and thus improving scale.
Staking on Aptos
As a Proof-of-Stake blockchain, $APT tokens come with a built-in utility, namely that of having voting power when it is staked. Voting power refers to validators agreeing, i.e. coming to a consensus regarding the ordering and execution of transactions before they are persistently stored on the ledger. Staking describes a kind of like locking the tokens into an escrow account as a sort of security deposit/collateral so that one can act on the voting power contained in the token.
Like in most PoS networks, validators voting power is proportional to stake (i.e. ‘tokens in escrow’) associated with it. However, there is a maximum amount of stake, that caps the rewards the respective validator can earn. This is set in place to disincentivize validators from accumulating an unhealthy amount of stake and as a result centralizing voting power.
Staking in Aptos works via three different ‘personas’: the owner, the operator, and the voter. One entity can be more than one persona at the same time.
- The owner owns the funds (duh!) and can add, unlock, and withdraw funds as well as change their validator. Rewards are deposited to the owner account and automatically compounded.
- The node operator performs the validation function and is assigned by the owner. The operator can enter or leave the validator set and change their own consensus key in case it gets compromised. The operator can not move any funds for which it does not have the ‘OwnerCapability’, i.e. funds for which it is not the owner.
- The voter is designated by the owner and participates in governance through their voter key.
Validators take turns as leaders that propose new blocks, which are then voted on by the network of validators to be finalized. For doing so, leaders earn rewards at the end of every epoch. Leaders are selected according to the deterministic formula (based on validator reputation determined by the validators’ performance + voting behavior + stake).
Epochs are time intervals on Aptos that dictate certain events, e.g. reward payment, validators entering/leaving the set, lock-up periods, etc. An epoch on Aptos is currently 2h (7200 sec) long.
Lock-up periods, i.e. the duration of stake ‘sitting in the escrow’, are automatically renewed every epoch unless otherwise declared by the owner.
Parameters such as minimum & maximum stake, duration of epochs, lock-up periods, etc. are all subject to on-chain governance and can be changed through proposals by the Aptos community.
Note: Precise, detailed information on certain staking parameters on Aptos is still scarce and we expect more to be available soon — we will update this section accordingly.
Learn “How to set up your Martian wallet & stake on Aptos via Tortuga Finance” with our step-by-step guide.
[updated 18. Oct. 2022]:
According to the Aptos Tokenomics post, the maximum reward rate starts at 7% and declines by 1.5% annually until it reaches a floor of 3.25%. The reward rate is evaluated every epoch.
Ok, now that you know what Aptos is, how it works (roughly), and what makes it special, a few words on why we are supporting the network.
Why we are backing Aptos
We are a firm believer in a multi-chain world in which a few blockchains power the internet of tomorrow. We set out to support networks, which we believe are among this set of blockchains. This inherently means that we are making bets on certain networks, one such being Aptos. There are several reasons why we are making this bet — let’s go over some.
Technology & second mover advantages
As you read above, Aptos brings many technological novelties to the table and their clear focus on mass adoption by improving UX through safety and performance resonates a lot with us. While we invest across the entire web3 stack, we have a past focus and proven track record of base layer investments. We believe Aptos is capable of carving out its fair market share amongst our portfolio networks.
They are taking a similar approach to Solana on many vectors, first and foremost, being a monolithic chain and leveraging the parallel processing of transactions but also by starting via the ‘non-EVM’ path. This proved to be very successful for Solana as their focus on Rust drew in a new, highly driven set of developers. Solana paved the way on many fronts and Aptos can draw from that. For instance, right from the start, there is already more of an ecosystem in place as there was for Solana in March 2020 — you already have different primitives such as wallets (Martian, Pontem, Petra), on-chain order books (e.g. Econia), NFT marketplaces (Topaz), or liquid staking (e.g. Tortuga) in place.
While some already dub Aptos as a ‘Solana killer’, we believe that they do not necessarily need to cannibalize each other, quite the opposite. We believe that both ecosystems can cross-pollinate due to their focus on low-level, very similar, programming languages. This could lead to an influx as well as education of new and experienced developers alike — a rising tide lifts all boats. Furthermore, both networks can learn from each other, carve out different niches, and serve different end users while bridges between them further reinforce synergies.
Team & backers
Spinning up a new layer-one blockchain and growing an thriving, self-sustaining ecosystem around it, are quite the tasks, especially in the current market and macro situation.
However, the team behind Aptos is well-equipped to surf these challenging waters. Parts of their team and both founders have prior experience working together during their time at Novi, a crypto platform, which is part of Meta (formerly Facebook). Novi focused on infrastructure and tooling for Diem (formerly Libra … this is getting out of hand).
Their co-founders, Mo Shaikh (link) and Avery Ching (link) draw from a rich background of experiences. Avery Ching (CTO), is a seasoned software engineer and worked as a Principal Software engineer at Facebook prior to Novi. He’s a long-time Apache Software Foundation member and has years of open-source experience as Apache Giraph VP. Mo Shaikh has a consultancy and TradFi background through his time at BCG, BlackRock, and the Royal Bank of Canada. After his time as director of strategy for ConsenSys, he worked on strategic partnerships at Novi. As a founder of a blockchain-based platform for fractional real estate investing, he has prior entrepreneurial experience in blockchain.
Their team of almost 80 people is comprised of experienced crypto product, BD, marketing & ecosystem growth managers as well as cryptography researchers and experienced developers. They are well networked to experienced providers of blockchain primitives such as infrastructure operators, wallet providers, and exchanges.
Through their investors, the team was not only able to raise a substantial war chest of 350MM USD. With renowned backers such as Multicoin, a16z, Jump Crypto, Binance, and FTX, the team has a far-reaching and fruitful network to draw from.
After years of prior development work, the network was stress-tested in three incentivized testnets and we successfully participated in AIT-3 next to several other experienced node providers. Our hands-on experience with the network convinced us even further of the teams’ capabilities, a strong, early network of supporters, as well as a promising ecosystem in its’ infancy. We look forward to being part of the Aptos ecosystem to help with the acceleration and adoption of web3!
Useful links, tools, and resources
- Martian: https://martianwallet.xyz/
- Pontem: https://pontem.network/pontem-wallet
- Petra: https://petra.app/