Parallel chain model
by Iurii Shyshatskyi, Chief Scientist at Jax.Network
Introduction
In the past four years, many notable projects have proposed their blockchain scalability solutions based on sharding. However, there is no common vision of how sharding should be implemented.
It is a common practice that founders of blockchain protocols assert superior scalability properties to their solutions using the following plot. When it is needed to show a superior throughput of their solution, they represent shards as independent units which can process transactions in parallel.
However, this parallel chain model is often an oversimplification. In order to work properly, every blockchain sharding protocol always leaves behind some link between chains so that they can interact and function as a whole. Obviously, the way links are set has a tremendous impact on the performance and other properties of the network.
In this post, we will discuss common cross-chain links native to PoS sharding architecture.
Throughout this paper, we say that there is a link between two chains whenever the validity of a block on the first chain relies on the data retrieved from the second chain.
Parallel chain model in media
Small little-known websites focused on cryptocurrencies introduce themselves as educators and entertainers who want to take up community awareness about complex concepts, promising cryptocurrency projects and their eye-catching ideas. In theory, this activity should stimulate adoption, however, in reality, it misleads people.
The parallel chain model is often discussed in cryptocurrency media resources which attempt to explain complex protocols based on Proof of Stake (PoS) sharding for dummies. The audience of these media outlets is eager to get an insight on scalability and transaction throughput of proposed solutions.
For example, in this article, the author asserts that with the launch of 64 shards, Ethereum 2.0 will have 64 times more throughput than Ethereum 1.0.
The problem here is that the blockchain community doesn’t have a common and rigorous way to define and estimate blockchain scalability. Blockchain geeks rely on an informal definition of scalability introduced by V.Buterin. Moreover, Buterin’s approach to quantifying scalability lacks rigor. It maltreats mathematical apparatus and misrepresents the concept of blockchain scalability. It’s not a surprise, the overwhelming majority of people are confused by it.
In order to fill the gap in their understanding, people use second-tier crypto-media that publish misleading explanations. As a result, many crypto-enthusiasts get confused and believe that a PoS network with N shards is N times more efficient.
Spokesmen of PoS projects stir up this misconception by throwing bold estimates that PoS sharding could bring “1000x efficiency improvement” to certain blockchain networks.
Apparently, there is no other way to introduce and explain the concept of sharding without referring to shards as parallel chains. However, in the case of PoS sharding, this approach is misleading.
Chain cross-links in PoS sharding
The fundamental problem of sharding solutions is the design of cross-shard transactions. As a matter of fact, their presence in the protocol testifies that links between shards are too strong. Verification of such a transaction often requires data from other shards. It may be not obvious, but this style of sharding never gives any benefits in terms of network traffic.
The design of cross-chain transactions is a typical issue of Proof-of-Stake sharding proposals. However, it’s not the only thing that heavily binds chains. The Proof-of-Stake consensus often requires complex, fragile and resource-intensive mechanisms for managing validator committees. PoS networks often have one special chain that stores records about past and prospective validators. As a result all other chains heavily rely on the state of this chain. This special chain is often called the beacon chain.
This data dependence works in the backward direction too: the correctness of account balances on the beacon chain relies on the shard chain data. Validators often get subsidies based on their participation in validation of shard blocks. The fundamental vulnerabilities of the PoS consensus require penalizing of validators who misbehave. However, this process is not as straightforward as it might sound since there is no central authority that can impose penalties. In the decentralized network, only peers can impose penalties by recording them into the blockchain. In order to verify account balances to the beacon chain in a trustless fashion, they need to know the state of every shard chain.
An illustrative example of the penalty is the Ethereum 2.0 protocol where the stake of the validator could get slashed if he goes offline too often or struggles to complete his assignment timely. As a result, validators in Ethereum 2.0 are forced to perfectly synchronize their clocks, validate block candidates in a due time and timely share data with their peers despite block propagation delays. In contrast, in Bitcoin, there is an asynchronous consensus protocol so that nodes might have inaccurate clock times and can go offline for whatever reason without breaking the flow of the protocol and without a risk of getting slashed.
It’s clear that aforementioned issues contaminate the economics of full nodes in sharded PoS networks. A minor stakeholder can expect only a minor income from staking which is proportional to the size of his small stake. However, if he wants to run a full node in the PoS network, he is forced to manage the network traffic from all shard chains. Therefore he is about to have expenses comparable to expenses of his wealthy peers. It’s bold to assume that his income from staking would cover his expenses on running a full node.
Conclusions
PoS sharding architecture deserves a significant portion of critics. However, an even bigger portion of critics should be forwarded to crypto media outlets. These online resources have become broadcasters of false, inaccurate, and misleading information.
Unfortunately, this commercial content often proliferates into academic research written by students and recent graduates. This substantial group of authors is used to work with well-established and thoroughly verified theories. Unfortunately, blockchain network theory is not mature. It’s infiltrated with toxic ideas and short-sighted vision. Working in this field requires a high level of critical thinking from researchers, crypto-enthusiasts and observers.
One might get dispirited and lose confidence in the future of blockchain sharding, scaling of blockchain technology and its further adoption. Luckily, the Jax.Network team has developed an alternative design of blockchain sharding based on the Proof-of-Work consensus. Shard chains in Jax.Network preserve a high level of independence. As a result, Jax.Network enjoys all efficiency benefits suggested by the parallel chain model. Chains in Jax.Network have only some limited linkage which we will discuss in our next post.
Don’t miss any announcement by subscribing to our social media accounts:
Official Website: https://jax.network
Twitter: https://twitter.com/CommunityJax
Facebook: https://www.facebook.com/network.jax/
Telegram Channel: https://t.me/jax_network
Official Group: https://t.me/jax_chat
