Scalability

Spark’s design enables it to be nearly infinitely scalable with minimal computational costs. This is unlocked through the lack of global consensus or propogation. Validator nodes (SOs) only need to gather for signing at moment of transfer. Transactions can be processed and settled independently of other transactions, making Spark fully parallelizable.

SOs themselves can be scaled horizontally allowing them to process and sign more transactions as they scale.

The goal of Spark is to be able to handle billions of concurrent users, and to do so with minimal fees and instant finality.