This technical architecture diagram shared by @ solayer_1abs explains the implementation logic (state synchronization aspect) of its million level TPS very well. Let me briefly explain my understanding: 1) This is essentially a sharded turbine architecture, which is a relatively high latency data broadcasting mechanism that achieves cross regional low latency transmission through dark fibers 1. This structure solves the bottleneck problem of single point bandwidth by implementing shard replication at various levels, and the system can achieve geometric level bandwidth expansion. Whenever data propagates through a new level (tree diffusion network), the total available bandwidth increases exponentially rather than linearly as the data is dispersed to more nodes; 2. It also combines compression processing and regional distribution to optimize network resource utilization. Its different data replicas are distributed in different environments, including cloud service providers such as AWS and Latitude, as well as personal environments such as Garage. These two points are specially designed, consistent with the design ideas of many high-performance layer1 public chains, except that Solayer adopts a larger scale, more layers, and finer granularity to support extremely high TPS requirements. 2) The PCDN - Distributed CDN Distribution Network Model is essentially an innovative mechanism for reducing infrastructure costs: 1. Decentralized storage and transmission: Traditional data centers have extremely high construction and maintenance costs, while PCDN allows ordinary users to contribute their idle storage space and network bandwidth, forming a "distributed cloud that everyone can participate in"; 2. Hierarchical service strategy: As can be seen from the graph, the system divides users into different roles, and verification nodes that require high response speed are located in professional environments, while ordinary users can contribute resources and receive corresponding rewards through PCDN; 3. Adaptive distribution: PCDN can intelligently identify user needs. For users who do not require real-time data, the system will provide services from the nearest or least loaded nodes, without having to connect to the central server every time, greatly reducing the pressure on the core network. above. We see that many existing projects encounter not computational bottlenecks but state synchronization issues after TPS exceeds 100000. Especially under high volatility network loads, can the state consistency of all validation nodes be maintained. Therefore, the core pain point that Solayer needs to solve is more of a state synchronization problem, which determines the actual effective TPS capability of a public chain. This technological architecture is not without challenges, as experience with Solana chains shows that this architecture, which heavily relies on physical infrastructure such as dark fiber, also brings challenges in decentralization and deployment costs. The introduction of the PCDN model can undoubtedly alleviate these problems in theory, but its performance in practical large-scale deployment environments still needs further verification.