|
Due to the "L2 scaling," which results in a poor user experience and a suboptimal token economy due to the inability to control costs within the L1 range, and the fragmentation that disrupts the user experience. Overall scaling methods (parallelization and sharding) will not fall victim to these fatal flaws, as it remains a cohesive whole.
Parallelization is inevitable because not multi-threading all client software is very foolish. All modern CPUs have multiple cores, yet blockchains like Ethereum (ETH) and Bitcoin (BTC) still process transactions sequentially. This leads to the inefficient utilization of most validator hardware, which is a significant waste!
The same applies to sharded chains, as each shard should undergo parallelization to maximize the capacity of an individual shard. Sharding is a logical development from parallelization. By scaling concurrency from multiple cores to now distributing workload among multiple computers, this breaks the previous scalability records.
Sharded systems are now able to achieve 100k+ TPS, with a theoretical limit approaching 1M+ TPS while maintaining relatively low node requirements. This is the approach to solving the blockchain trilemma through sharding.
Traditional blockchain designs face the blockchain trilemma. Because at some point, node requirements become so high that they threaten decentralization. Since all nodes must validate all global state changes, it is fundamentally not scalable. Sharding technology addresses this issue.
Unlike traditional designs, sharded chains scale with usage. Non-sharded chains, on the other hand, always run into bottlenecks eventually. As a sharded chain gains more adoption and validators through increased usage, it can initiate a new shard. In other words, sharding makes scaling linear!
In contrast, other blockchains exhibit quadratic scaling, meaning that as the network grows, node requirements increase until reaching physical limits. This is because, compared to what we can achieve through computer networks, we can only process so much within a single silicon chip. |
|