What Is Harmony Trying to Achieve?

Higher transaction throughput is what should set Harmony apart from the likes of Ethereum and other blockchain solutions which are forced to achieve performance gains by sacrificing other features. Harmony developers describe competing solutions as unable to resolve scalability issues or provide support for applications which require high throughput performance, such as with gaming or decentralized exchanges. Similarly, blockchains such as EOS or IOTA tried replacing consensus models and introducing new tech, such as directed acyclic graph (DAG). All of these came at the expense of security and/or decentralization which Harmony aims to preserve by creating shards (groups) of validators that would be able to process transactions simultaneously. Based on this, the total transaction throughput should increase in a linear manner and in parallel with the growth in the number of shards. In September 2018, Harmony’s testnet managed to achieve 118,000 TPS with some 44,000 nodes, with the hope to close the gap to the Visa’s 2,000 TPS on a daily basis.

Harmony’s consensus protocol goes for speed and energy efficiency. Much of the Harmony’s scalability and throughput promises rest on the ability of its Fast Byzantine Fault Tolerant protocol (FBFT) to employ parallel transaction processing to scale with the size of the network and effectively tackle its connection latency. Its network topology is designed to enable for faster consensus reaching and message exchange. At the same time, Harmony features a kernel designed to run its protocol in a manner which allows a wider range of devices to participate in the consensus building, thus strengthening its decentralization. The deep sharding process itself relies on adaptive proof-of-stake model based on distributed randomness generation (DRG) procedure which is described as secure, easily verifiable and scalable.