Blockchain

The Bank for International Settlements (BIS) has recently announced the latest progress on Project Agorá, showcasing how tokenization technology is being applied to cross-border wholesale payment scenarios. This initiative brings together seven central banks and over 40 financial institutions. Leveraging a dual-layer blockchain framework and atomic settlement mechanisms, Project Agorá enables cross-border funds to be settled within seconds once liquidity is secured, while simultaneously mitigating settlement and credit risks. Project Agorá is also recognized as a significant experiment in the digital transformation of global financial infrastructure.

As global demand for cross-border payments continues to rise, improving payment efficiency, lowering settlement costs, and maintaining financial security have become critical issues for central banks and financial institutions around the world. Project Agorá, spearheaded by the Bank for International Settlements (BIS) in partnership with several central banks and financial entities, is a research initiative focused on cross-border payments. The project seeks to reimagine wholesale cross-border payment workflows using technologies like tokenized finance, shared ledgers, and atomic settlement.

Global financial infrastructure is steadily advancing toward tokenization and digitization. Cross-border payments are transitioning from traditional bank settlement models to innovative frameworks that incorporate blockchain technology and shared ledgers. Project Agorá, an important research initiative spearheaded by the Bank for International Settlements (BIS) alongside several central banks and financial institutions, is dedicated not only to enhancing payment efficiency, but also to building a next-generation cross-border payments ecosystem that upholds both financial stability and regulatory compliance.

KAS was launched fairly, with no pre-mining and no ICO, and all tokens are released through competitive KHeavyHash farming. The total supply is limited to about 2.87 billion tokens. The issuance curve includes a pre-deflationary phase and a chromatic phase, with block rewards smoothly decreasing based on DAA scores. KHeavyHash utilizes Keccak hash matrix multiplication as its core, with miners and RustyKaspa full nodes collaboratively maintaining blockDAG security.

The primary distinction between Kaspa (KAS) and established PoW blockchains like Litecoin (LTC) and Monero (XMR) is their ledger architecture. Kaspa employs a blockDAG alongside the GHOSTDAG consensus mechanism to achieve parallel, high-frequency block generation. In contrast, Litecoin and Monero retain the conventional single-chain linear block structure, with each blockchain adopting unique approaches to payment efficiency and on-chain privacy.

Kaspa (KAS) is a Layer 1 public blockchain utilizing Proof of Work (PoW), replacing the conventional single-chain architecture with blockDAG technology. Leveraging the GHOSTDAG consensus protocol, Kaspa organizes parallel blocks into an orderly ledger, aiming for a block production rate of about 10 blocks per second. The network was launched fairly, with no pre-mining or undisclosed allocations. It employs the KHeavyHash mining algorithm, and its node implementation is RustyKaspa.

GHOSTDAG consensus serves as the primary ordering engine in Kaspa (KAS) PoW blockDAG architecture. This allows miners to broadcast multiple valid blocks simultaneously, while GHOSTDAG applies Blue/Red classification and k-cluster rules to transform the parallel block graph into a globally consistent, sequential ledger.

The fundamental distinction between Kaspa (KAS) and Bitcoin (BTC) is in their ledger architectures: Bitcoin employs a single-chain, linear block structure to chronicle transaction history, whereas Kaspa leverages a blockDAG, enabling parallel block generation and establishing a global order via the GHOSTDAG consensus mechanism. Although both systems utilize Proof of Work (PoW), they differ in terms of data architecture, block generation speed, orphan block management, and farming algorithm.

Nesa's AI inference process is the end-to-end execution workflow for an AI request: from submission and task scheduling to distributed inference, result verification, and final output delivery. By integrating the MetaInf scheduling system, private inference technology, and verification mechanisms, the process enhances data privacy and result trustworthiness while performing AI inference.

Verifiable AI is a technical mechanism that authenticates the actual execution of AI reasoning, ensures the trustworthiness of output results, and enables independent verification. Nesa embeds Verifiable AI as a core network capability, using cryptographic proofs, distributed execution, and result verification to not only complete AI inference computations but also to prove that the reasoning process meets expectations, thereby strengthening developers' confidence in AI-generated outputs.

Nesa (NES) is an AI infrastructure network designed for privacy-preserving, verifiable computing and decentralized execution. Leveraging Equivariant Encryption (EE), HSS-EE, and a distributed inference architecture, it allows AI models to perform inference tasks without exposing data or model content.

Brevis is a verifiable computation platform driven by zero-knowledge proofs (ZK), referred to as "The Infinite Compute Layer" for Web3. Its core approach is to offload resource-intensive computations off-chain, enabling the blockchain to simply verify a succinct proof on-chain.

Pico zkVM is the open-source modular Zero-Knowledge Virtual Machine (zkVM) from Brevis (BREV). It allows developers to write any computational logic in Rust and produce a Zero-Knowledge Proof (ZK) verifying that the computation was executed correctly.

The ZK data coprocessor eliminates smart contracts’ inability to access historical data by enabling off-chain retrieval of historical or cross-chain data from archive nodes, performing computations, and returning a zero-knowledge proof that confirms both the result and the authenticity and correctness of the data. This proof can be verified by the contract on-chain in milliseconds. The data flow involves four stages: application request, off-chain computation, ZK proof generation, and on-chain verification and result reception, allowing for reliable conclusions without the need to replay all data.

Oracles primarily bring external, off-chain data onto the blockchain, with their credibility relying on nodes and economic incentives. ZK co-processors such as Brevis focus on verifiable computation of existing on-chain historical data, leveraging zero-knowledge proofs to allow direct on-chain validation of results. These two technologies occupy opposite sides of the blockchain data flow, providing complementary roles in data direction, trust models, and computational capabilities. Their boundaries are not strictly defined; in real-world scenarios, collaborative hybrid approaches are commonly adopted.