Connor Daly 🐈(4x)πŸ”Ί
Connor Daly 🐈(4x)πŸ”Ί

@das_connor

35 Tweets 69 reads Nov 10, 2021
Alright, time for a mega thread.🧡
I call this one:
Subnets or: How Avalanche Plans to Take Over the World.
To understand the power of subnets, we need to talk about blockchain scalability.
Without sacrificing decentralization, there are two ways to expand blockchain capacity: vertically or horizontally.
In short, faster chains or more chains.
1. Vertical scaling.
Process transactions faster.
The current bottleneck here is storing smart contract call data. Each new block updates the client's internal data structures. These updates require several non-contiguous writes to disk.
Across ecosystems, people are making incremental improvements.
Most come by creating more efficient data structures or enforcing beefier hardware requirements.
These gains are gradual and likely won't lead to 1000x returns.
2. Horizontal scaling.
Increase capacity by executing multiple blockchains in parallel.
This is where you can have almost unlimited throughput gains.
For example, let's say we have a blockchain that can do 100 transactions per second (TPS).
If we run five of them in parallel, our system now supports 500 TPS.
But of course, there's a catch.
For chains to operate in parallel, they need to be independent.
That means that individual chains need to be self-contained ecosystems. If you want interoperability, you need to add a cross-chain communication layer.
Despite the benefits of horizontal scaling, it's not a common approach.
Almost every L1 project runs a single chain.
Subnets are Avalanche's form of horizontal scaling.
But we don't plan on copy and pasting the same blockchain over and over again.
Instead, we're creating a diverse multi-chain ecosystem.
@avalancheavax already contains three primary blockchains.
β€’ The C-Chain is an EVM chain for smart contracts
β€’ The X-Chain is a UTXO-based chain for payments
β€’ The P-Chain is a custom chain for platform governance and validator setup
Currently, most Avalanche traffic runs on the C-Chain. That's where DeFi and NFTs currently thrive.
But if you've ever staked or delegated your AVAX, you've used the P-Chain.
Assets can already bridge across chains. Native AVAX can move across all three.
ERC-20s can move from the X-Chain to the C-Chain and back.
As builders create new subnets, this list of chains will grow.
One day yours too might be on this illustrious list.
Subnets require two things to build:
1. A virtual machine: How to process transacations
2. A validator set: Who can process transactions
A virtual machine defines the rules for processing transactions.
Everyone's favorite example: the Ethereum Virtual Machine (or EVM).
But that's not the only one!
Others are the Bitcoin Script VM, Cardano's UTXO model, Solana's transaction engine, etc.
An Avalanche subnet can run any of these.
Forks can even bootstrap with old chain state.
For example, you could create a Bitcoin subnet and airdrop AVAX-Bitcoin to all existing BTC holders.
While the EVM is the present, I'm not convinced its the future.
Despite HUGE network effects, we've seen changes in every layer of the software stack.
Operating systems, programming languages, processor architectures, internet standards.
Why should blockchains be any different?
Eventually, someone will create a VM that's superior to the EVM, and the world will adopt it.
That VM can launch on Avalanche without the protocol needing to reinvent itself.
This future proofs the network.
Next up, the validator set.
Validators take part in consensus and are the keepers of the network.
By tweaking the validator set, protocol designers determine the visibility and accessibility of the subnet.
Legal validator behavior determines who can issue transactions as well as who can see the chain and download its blocks.
This means that a subnet can be permissionless and open to the public, private and only visible to a select group, or something in between.
A more complicated example:
Let's say the US government wants to create a digital dollar subnet.
They can build a chain that's publicly visible but restricts validator participation to US citizens.
Ultimately, subnets will support public goods AND enterprise use cases.
But why would anyone choose to build a subnet?
If you've gone through the work of building a custom VM, you might as well launch your own L1, right?
Well, here's a few benefits to building a subnet instead.
1. Abstracted low-level complexity
2. Instant access to liquidity
3. Choice of gas token
The Avalanche protocol provides a consensus engine and manages communications between nodes.
This simplifies what you need to build.
AND you get the best part of Avalanche, Avalanche consensus, for free.
This means high TPS and fast time to finality by default.
At Ava Labs, we know a thing or two about bridging.
That's why subnets will also provide frameworks for interoperability and cross-chain communication.
Subnets work best when you can seamlessly import and export assets across chains.
Building on Avalanche will connect you to existing ecosystem liquidity from day one.
You can also launch your own gas token. When you create your own subnet, you can define what token you use to power it.
It doesn't need to be AVAX. Transactions could even be free.
This gives builders an opportunity to raise capital to support their project.
For more Avalanche explainers, check out my pinned tweet.
I write about how to get started building in crypto and target all skill levels.

Loading suggestions...