The Nervos Network is a two-layer blockchain. Layer 1 is the Nervos CKB (Common Knowledge Base) which uses a proof of work protocol. This base layer allows any crypto asset to be stored with the security, immutability and permissionless nature of Bitcoin. On top of that is layer 2 – which enables users to create their own blockchains, protocols and integrations.
Today we talk to Jan Xie, Chief Architect of Nervos, to understand the logic behind this innovative architecture and the potential it holds for growing DeFi.
Rochelle: What challenges do Bitcoin and Ethereum face that you fix?
Jan: A common challenge for both is scalability and our answer to that is a layered architecture.
For Bitcoin, a fundamental and often overlooked problem is sustainability, or in other words: “will it last 100 years?”. Bitcoin’s issuance curve (halvening every-4-years) and transaction fee model means miners will receive less and less rewards for protecting the network unless Bitcoin’s price doubles every 4 years or transaction fees increase.
The first situation is very uncertain, if not impossible. The second will create a vicious feedback loop in which high transaction fees reduce on-chain activity, while less on-chain activity requires even higher fees.
At the moment, Bitcoin’s “store of value” and 2nd layer scalability goals reduce first layer transaction activities and therefore reduce fees for paying miners and securing the network.
Ethereum faces a different challenge. We call it the “heavy asset problem”. Ethereum pays miners ETH to incentivize consensus and keep the network secure. However, Ethereum is a store for many different tokens, who benefit from Ethereum’s security but do not contribute to it. Instead it is the ETH holders who have to pay the on-going storage cost (through creation of ETH to pay miners).
In fact, the more successful these non-ETH tokens are (with the growth of DeFi for instance), the more lucrative and attractive Ethereum is to attackers. Meanwhile the security of the network has not increased proportional to the total value of all tokens stored on the network.
The economic model of Ethereum contradicts the ecosystem it enables. Without sustainability, a blockchain will hit its growth ceiling eventually. The more successful and faster it grows, the sooner it will meet its inherent bottleneck.
We think these are the most important problems to be solved to create a truly decentralized and sustainable blockchain network – and they can only be solved by a complete new design.
Our solution is the Cell Model, which is a generalization of Bitcoin’s UTXO model. It is designed for the storing of multi-assets and a layered architecture, in which miners get rewards proportional to both transaction activity and the length of time tokens are held. We make Nervos very simple and abstract so it can evolve, together with a correct economic model we believe will outlive ourselves.
Rochelle: Another approach to solving the scalability trilemma has been to use a different protocol than Bitcoin’s proof of work, such as proof of stake. Why did you decide to build a two-layer blockchain instead?
Jan: The main strength of a system is usually also its main weakness. And for blockchains, both are centered around consensus. We trust blockchains because anyone can participate to verify transactions. But as a result, this restricts scalability because each vote and transaction needs to be broadcasted and processed by everyone.
If we look at both ends of the scalability spectrum, on one end we have a centralized system which can scale well but where only one entity can verify the results. On the other end we have a decentralized system in which everyone can verify everything, but which can’t scale as a result.
For a blockchain to be a store of value accessible to everyone in the world, truly verifiable and trusted by all – the only possible choice is to let everyone verify everything. This makes layer 2 solutions the only option to achieve scalability.
This is our design principle at Nervos: layer 1 for security, finality and decentralization, and layer 2 for scalability.
Rochelle: One major issue with Bitcoin is that it is very centralized, due to the cost of mining. How does Nervos ensure decentralization of its network?
Jan: I think the most important feature of Bitcoin is the fact that it is permissionless. No matter who controls the token supply or hashrate, the system is always open to newcomers. There’re capital requirements for participants of course, but the process itself is open and fair.
It doesn’t really matter if Bitcoin is “70% or 20% centralized” at any time point, as long as it’s permissionless it will be eventually decentralized. It’s a dynamic system, we should look at it with a telescope of time. Electricity generation will become more decentralized as technology advances. I also believe energy resources such as solar, wind and hydropower are more decentralized than resources like pre distributed tokens or reputation (used in PoS).
Besides that, we need to make ASICs as accessible as possible so anyone is able to mine. That’s the reason we designed a new hash function called Eaglesong which focuses on security, novelty and simplicity. Novelty means there is no existing hardware optimized for Nervos before mainnet launch, so everyone starts from the same point. Meanwhile simplicity of design means hardware manufacturing is more accessible and can be decentralized more easily.
Based on what we have seen so far, this strategy is working. The Nervos hash rate has increased over 100x in 6 months since its launch, with 3~4 ASICs manufacturers competing with each other. A PoW network is not born secure, it’s only secure with a successful bootstrap and I think we have made it.
Rochelle: The DeFi ecosystem has largely been populated by Ethereum-based protocols as a result of its early beginnings at the tail end of 2018 with the launch of MakerDAO. How do you see Nervos impacting this ecosystem?
Jan: New ideas, new developers, new users.
The shape of the DeFi ecosystem today is too affected by Ethereum: tokens are pooled together in ERC20 contracts; transactions must be hashed by Keccak and signed by Secp256k1; transaction fees must be paid in ETH, even when the transaction fails; front running is too easy, etc.
DeFi on Nervos will be very different. For one, using different signing algorithms to verify transactions means DeFi could be supported by a wider range of devices. Second, accepting non-native tokens would make DeFi accessible to more users. Third, our transactions are deterministic which makes it harder for a miner to front-run them.
Because the design of the Nervos blockchain pushes many functions to the upper layer, DeFi on Nervos will naturally rely more on open transactions (“layer 1.5”) and layer 2 protocols. I hope we can bring some new ideas to this space.
A prosperous DeFi ecosystem needs builders. On Nervos, developers are not limited to Solidity, and you don’t need to learn a new language to develop smart contracts. This freedom of choice lowers the entry barrier for developers outside the blockchain community and could bring more developers into the nascent DeFi ecosystem.
Rochelle: Your platform allows new blockchains to be created on top of the Nervos CKB. Has this architecture allowed for more DeFi use cases to emerge, where regulations had so far limited adoption, for instance in China?
Jan: Definitely. One of the benefits of a layered architecture is the fundamental layer can be kept neutral, while allowing the upper layers to customize for different regulation conditions in different countries and regions. In China the majority of blockchain applications are built on permissioned/consortium blockchains for regulation reasons. The permissioned blockchain market is growing fast and we’re working closely with many players to remain at the forefront of it.
For instance, we’ve partnered with Huobi chain to expand the DeFi market in China. It may take tremendous effort and understanding of the market to make DeFi available to Chinese users, but we won’t stop exploring the possibilities that can make it happen.