An Introduction to the Avalanche Blockchain
Introduction
Avalanche is a relatively new blockchain that launched in September of 2020. It aims to be the fastest smart contracts platform available with a new consensus mechanism that solves the scalability issues other blockchains have while providing more flexibility in use cases.
While blockchains have already disrupted the financial industry and will continue to do so, DeFi on blockchains like Ethereum remain slow and expensive to use due to high gas fees and low throughput. In addition, Bitcoin’s high security, but low functionality scripting language is not built to support the plethora of use cases that will be needed in a vibrant economic system. Avalanche strives to be the solution to building blockchain infrastructure for three main use cases: (1) application specific blockchains, (2) scalable and decentralized applications, and (3) a variety of smart assets.
Contrary to the common understanding of the term “blockchain”, where all transactions are validated by all validators, Avalanche is made up of a collection of blockchains that are validated by subnets of validators. Aside from the requirement to help validate the main chain, nodes are able to choose the chains they want to help secure. In the long run, the Avalanche team hopes this will encourage more nodes to participate since they are able to opt-in to securing the networks they care about, and this option lessens the storage and computational requirements of nodes that help secure the network.
Team Background
The team at Ava Labs is led by some industry experts in the fields of computer science, cryptography, blockchain, and asset management:
- Emin Gun Sirer, CEO: Sirer was an associate professor at Cornell University for 20 years with a background in computer science and peer-to-peer systems before focusing on Avalanche full time in August of 2021.
- Kevin Sekniqi, COO & Chief Protocol Architect: Sekniqi is a PhD student of philosophy at the University of Cornell with experience working on homomorphic encryption research at the City College of New York
- Chris Lavery, CFO: Lavery was VP of Finance and Operations at Blockchain.com before he joined Ava Labs
- John Wu, President: Wu started as an analyst and portfolio manager for various asset management firms. He then founded and continues to serve as the CEO of Sego, an investment firm focussed on technology, before he joined Ava Labs as President
- Patrick O’Grady, VP of Engineering: O’Grady was previously a crypto engineer at Coinbase before leaving to work on Avalanche
- Lydia Chiu, VP of Business Development: Chiu was a senior analyst at Sego Capital for 12 years before joining the Ava Labs team
Technology & Network Architecture
The Avalanche network wants to give developers the tools to create highly scalable, customizable, and application-specific blockchains by incorporating a directed acyclic graph (DAG) where the time-to-finality of blocks is a few seconds. Validators in the Avalanche ecosystem only validate the Primary Network and the subnets they want to participate in.
The Primary Network
There are three chains that form the Primary Network on Avalanche: the C-Chain, the X-Chain, and the P-Chain. These blockchains are built into the Avalanche platform and every validator is required to help validate these chains. The rest of the blockchains that exist on Avalanche are spun up by users of the network. Anyone can create a blockchain and gather a subnet of validators to help secure that particular chain.
The three default chains have their own purposes.
The C-Chain stands for the Contract chain and allows for the creation of Ethereum-compatible smart contracts. It implements the Snowman consensus protocol.
The X-Chain stands for the Exchange chain, which allows for the creation of assets on the Avalanche network. It implements the Avalanche consensus protocol.
The P-Chain stands for the Platform chain. This chain is used for keeping track of validators, active subnets, and enables the creation of new subnets. It implements the Snowman consensus protocol.
Subnets
A unique property that Avalanche brings to the blockchain space is the flexibility that comes with subnet-verified blockchains. As blockchains continue to struggle with finding solutions for scalability while maintaining decentralization, many of them are turning to various forms of sharding. Avalanche implements this from the get-go, making the hardware requirements for running a node less computationally intensive. Subnets give nodes, or validators, the option to pick and choose which blockchains they want to help secure and validate. As mentioned above, aside from the three default chains, nodes are able to opt-in on participating in a particular subnet that secures a specific blockchain in the Avalanche network.
Consensus Mechanism
The purpose of a consensus mechanism is to achieve agreement on the current state of the network. In some blockchains, this is done by choosing one miner or validator to add the next block. The selection process is what introduces economic dynamics that can, over time, favour the nodes with more resources. We have yet to see how this will affect blockchain networks over a long time horizon, but it is already evident that this jeopardizes the decentralized nature and spirit of blockchains. Block time limits also affect the scalability, cost, and overall adoption of blockchain technologies. With the fastest time-to-finality, Avalanche claims that its consensus mechanism design can solve scalability issues while remaining decentralized and avoiding the pitfall of allowing whales to gain unfair advantages.
Slush, Snowflake, and Snowball
In the Avalanche network, there are three iterations of consensus mechanisms that build on top of each other: Slush, Snowflake, and Snowball. These mechanisms mimic gossip protocols where validators randomly ask other validators in the network about whether a transaction is valid. After a certain number of random subset sampling, false transactions would be rejected and nodes are steered towards a common and correct outcome. This whole process happens very quickly, which is how Avalanche claims to have the fastest time-to-finality compared to all other blockchains. The philosophy of the snow family protocols is “metastability”, or to reach the correct conclusion about whether to accept or reject a transaction as quickly as possible.
Slush is the first step of the consensus mechanism progression. In the process of deciding whether to accept or reject transactions, every node randomly samples a subset of other nodes in the subnet. A node will change its decision based on the majority decision of each subset that it samples.
Snowflake is the second step of the progression. To prevent the slush random sampling process from going on forever, snowflake introduces a counter to every node as a form of memory. This counter increases every time the slush sampling comes to the same outcome. If a sampling produces the opposite outcome, the counter resets to zero. The counter serves as an indicator of the strength of conviction that a node has for its current decision.
Once again, to prevent the Snowflake process from going on forever, the Snowball step introduces a ceiling number to the counter. When the counter reaches this number, that node is locked on that outcome. For example, if a certain node samples ten different subsets of nodes and every single random sampling reaches a conclusion to accept the transaction, that particular node is locked in to ‘accept’. This introduces an aspect of finality to the process.
To become a validator, users must stake AVAX. Subnets, however, can use PoW or PoS, which is an interesting decision for the engineers to allow users to do. Unlike other blockchains, they also chose to not slash participants that misbehave, stating the decision as a prevention of malfunctioning hardware or software leading to a loss of coins. Due to the random sampling process, nodes that initially disagree with the majority will eventually be steered towards the right outcome so it is unclear whether misbehaving nodes need to be punished.
In many consensus mechanism designs, the chance to mint and collect minting rewards is proportional to computational power or existing funds. This creates economic dynamics where some nodes may have a competitive advantage because they have more resources to begin with. In the Avalanche network, node rewards are proportional to their uptime and response latency. Instead of proof-of-work or proof-of-stake, Avalanche uses proof-of-uptime and proof-of-responsiveness to prevent what they call “rich-get-richer compounding effects”. Nodes can begin to mint by staking and then participating in the consensus process.
Crypto-economics
Token supply and uses
The AVAX token has a capped supply of 720 million tokens. 360 million of these tokens were minted for investors at launch and have a vesting period from one to ten years. The remaining 360 million tokens are for staking rewards that will be released over time to validators. The minting rate of the AVAX token is dependent on factors like staking rewards, fees, and airdrops.
Holding the AVAX token allows users to participate in on-chain governance. However, only pre-determined parameters can be changed through token voting to make sure the platform stays relatively stable for users. Furthermore, governable parameters are subject to something called hysteresis, meaning there are limits to how much these parameters can be changed within certain time frames. If a parameter has undergone a drastic change based on votes, it is unlikely that it can be changed drastically again in the near future. This is to ensure that the Avalanche network can evolve over time, but remains relatively stable in short time horizons so the system is predictable and users can plan for these changes accordingly.
Unlike other blockchains where the fees are proportional to the congestion on the network and are not subject to changes, fees on Avalanche can be recalibrated at the end of a designated time period. This means that if the amount of activity on the network stays high for extended periods of time, reflecting more adoption of the network in general, the fee-function will lower fees.
Ecosystem Metrics
DeFi on Avalanche didn’t see much growth until September of 2021 when Olaf Carlson-Wee’s Polychain Capital and Su Zhu’s Three Arrows Capital co-led a $230 million dollar investment in the ecosystem. Since then, the TVL on Avalanche has skyrocketed to about $11.6 billion USD.
Alongside the massive increase in DeFi TVL, the number of unique wallet addresses and the number of daily transactions have both gone up by around 10x since September 2021.
Risks & Obstacles to Success
Avalanche is a heavily funded project. The fact that it has so much VC support can be a good sign since investors who have teams of cryptographers, researchers, analysts, and capital allocators are willing to believe in the project and bet on its potential. However, with VCs injecting so much capital into a blockchain and artificially creating momentum for its growth, it’s difficult to tell how many users and developers truly understand and are aligned with the network, versus how many are simply there for the hype and short term capital gains. On one hand, pouring capital into a blockchain to jumpstart its ecosystem is a growth strategy — no different from having a marketing budget for a traditional business. However, blockchains are globally distributed networks — they have an aspect of community and hive-mind network effects that can make or break blockchain success during rough market conditions. Long-term blockchain success does rely on informed users and developers who believe in the value proposition of the network and want to participate in its growth.
Furthermore, Avalanche introduced a new consensus mechanism that has not been battle-tested before. As a relatively new chain that launched into a market filled with vibrant layer 1’s that are all trying to plant their flag in the ground and increase their market share, Avalanche will have to make their value proposition very clear to attract developers who would rather build in a safer blockchain environment with a plethora of developer tooling to make their lives easier. In a space where forking code and making tiny improvements is very common, the novel consensus mechanism that Avalanche brings to the table, is both an advantage and a risk.
Final thoughts
By introducing a new family of consensus mechanisms, Avalanche is definitely venturing into the unknown. The project has a team with lots of technical expertise to build out the ecosystem it has envisioned. Similar to many other chains that are heavily funded, it will become more clear over time whether Avalanche has a truly dedicated community that can help it grow into one of the long-term players in the crypto space.
