BlockDAG technology, created by the Kaspa developers, is the next leap beyond legacy blockchain implementations. Kaspa’s pioneering blockDAG is a fully decentralized, intrinsically scalable solution that adheres to the highest security standards through a generalization of the Nakamoto Consensus model.
BlockDAGs resemble a directed acylic graph – a mathematical structure where vertices represent blocks, and edges reference child & parent blocks.
Where traditional blockchains appear as a simple linear structure, blockDAGs resemble a complex, interconnected web. Rather than chaining to singular blocks (as in normal blockchains), blockDAGs reference multiple predecessors & all tips of the graph structure, incorporating blocks from various branches instead of following a single, main branch.
Kaspa uses GHOSTDAG – an advanced blockDAG protocol that generalizes the Nakamoto Consensus model.
Kaspa is able to create up to 18 blocks in parallel, fully utilizing proof-of-work efforts by chaining all mined blocks within a directed acylic graph structure.
This parallelism prevents the underlying consensus layer from splintering due to extreme block speeds, and currently enables the largest throughput of any known proof-of-work network.
Kaspa currently propagates at one block per second, making it the fastest proof-of-work cryptocurrency.
Transactions submitted to the network typically receive their first confirmation in the following second. Full confirmations can take up to ten seconds, as a measure of security for the end-user.
Kaspa is propagated by a robust proof-of-work engine – a network of miners racing to solve blocks, every second. This time-tested consensus mechanism permits ultimate security.
Kaspa maintains consensus in a fully decentralized manner, as the high block throughput permits low variance, thereby incentivizing solo miners to operate their own nodes locally for low latency.
Kaspa is the result of over eight years of theoretical research & design into blockDAGs. Starting from GHOST, each implementation underwent numerous series of modifications & improvements, finalizing with the current advanced GHOSTDAG protocol which was translated from theory into a functioning permissionless ledger for Kaspa.
Written in 2013 by Kaspa visionaries Yonatan Sompolinsky & Aviv Zohar, the GHOST (Greedy Heaviest-Observed Sub-Tree) protocol was the precursor to blockDAGs. Although GHOST did not conceptualize usage of DAGs in its theory, it outlined the orphan block issues experienced by blockchains with high block rates, and proposed to solve them by indexing orphaned blocks in the deterministic mechanism that calculated the longest chain.
This novel protocol was cited in the 2014 Ethereum Whitepaper.
Written & published in 2016, the SPECTRE (Serialization of Proof-of-work Events: Confirming Transactions via Recursive Elections) protocol was the first implementation of blockDAGs in theory. It inherited the GHOST protocol’s orphan block mechanism, and generalized Nakamoto’s blockchain into a direct acyclic graph. This special design, in theory, enabled miners to create blocks concurrently and at a far higher rate.
Published in 2018, the PHANTOM protocol further extended on SPECTRE and presented a protocol that was capable of exceptionally large transactional throughput, through usage of blocks structured within a directed acyclic graph. It solved the core challenge of transactional ordering within a DAG protocol, and presented a full order on the blockDAG in such a manner that honest nodes could retain network consensus.
Published in 2021, the GHOSTDAG protocol was written by Yonatan Sompolinsky, Shai Wyborski, & Aviv Zohar, and is the most advanced implementation of blockDAGs. It extended on PHANTOM by integrating an efficient greedy algorithm to finding the maximum k-cluster subDAG (a variable used in determining the “well-connectedness” of a set of created blocks).
This paradigm, which generalized Nakamoto’s chain into a blockDAG, while avoiding the scalability-security tradeoff imposed by Nakamoto Consensus, is the underlying technology behind Kaspa, which launched around the same time as the publication of the GHOSTDAG paper.
Released on October 31, 2022 on the 14th anniversary of the Bitcoin whitepaper, the DAG KNIGHT Protocol – written by Yonatan Sompolinsky and Michael Sutton – is a parameterless generalization of the Nakamato Consensus methodology.
This novel protocol assumes no upper bound to network latency and allows the protocol to self-adjust to real-time latency. This in turn allows any protocol based off of DAG KNIGHT to be secure against any attacker with under 50% of network computational power while operating at extreme speeds & throughput.
DAG KNIGHT is a theoretical breakthrough, and is the first to theorize & prove a protocol can operate at a self-adjusting network latency. It is expected that Kaspa will receive the DAG KNIGHT upgrade some time in 2023.
Kaspa is currently undergoing an extensive core rewrite in RustLang. This performance-oriented programming language will transform the Kaspa codebase to its purest form – enabling record efficiency & industry-leading performance, documented, modularized code for future developers, and the groundwork for smart contracts.
The Rust rewrite will introduce unprecedented throughput, with initial estimates targeting 32 blocks per second (one block every ~31 milliseconds), with future projections of 100 blocks per second. Kaspa – after the Rust rewrite – will be the fastest & highest throughput decentralized network in the industry.
You can follow progress on the Rust rewrite below. It was completed in Q2 2023, and is currently undergoing public testing before being pushed to mainnet.