Cortex Blockchain
The first and currently the only public blockchain capable of executing on-chain AI
The current challenge to execute machine learning programs on the conventional blockchain is that the virtual machine is extremely inefficient when running any nontrivial machine learning models. Therefore, most people think it is impossible to run AI on the blockchain.
Cortex is an open-source, peer-to-peer, decentralized blockchain that supports Artificial Intelligence (AI) models to be uploaded and executed on a distributed network. Cortex achieves AI democratization by providing an open-source AI platform where AI models can be easily integrated into smart contracts to create AI-enhanced decentralized applications (DApps).
Cortex is an open-source, peer-to-peer, decentralized blockchain that supports Artificial Intelligence (AI) models to be uploaded and executed on a distributed network. Cortex achieves AI democratization by providing an open-source AI platform where AI models can be easily integrated into smart contracts to create AI-enhanced decentralized applications (DApps).
ZkMatrix
ZkMatrix is a Layer 2 Solution for EVM compatible blockchain powered by Cortex Labs. It utilizes zkRollup technology to aggregate batch transitions on the side chain into one transaction on the main chain, which aims to expand transaction per second(TPS) and significantly reduce the Endorphin fee consumed by each transaction.
ZkMatrix inherits the security and consensus aspects of the main chain and focuses on executing the transaction performance. In the event of a cyber attack or malfunction, users can still restore the state of the side chain based on the data in the main chain.
It is divided into three modules: ZkWallet(front-end), ZkNode(side chain), and ZkProver. The process is as follows:
ZkWallet sends a transaction to ZkNode for aggregation, and the status changes from "queued" to "committed".
ZkProver generates circuit proof and submits an aggregated transaction to the main chain, and the status change to "verified".
ZkProver utilizes GPU and CPU chips to shorten the time required to generate Proof and supports multiple Provers in parallel on ZkNode. ZkMatrix can support transactions up to 2000 TPS and the entire process takes just 5 seconds. After the data is sent back to the main chain, to achieve maximum security, it is estimated a wait of 60 seconds for the data to be actually stored in the main chain.
ZkMatrix inherits the security and consensus aspects of the main chain and focuses on executing the transaction performance. In the event of a cyber attack or malfunction, users can still restore the state of the side chain based on the data in the main chain.
It is divided into three modules: ZkWallet(front-end), ZkNode(side chain), and ZkProver. The process is as follows:
ZkWallet sends a transaction to ZkNode for aggregation, and the status changes from "queued" to "committed".
ZkProver generates circuit proof and submits an aggregated transaction to the main chain, and the status change to "verified".
ZkProver utilizes GPU and CPU chips to shorten the time required to generate Proof and supports multiple Provers in parallel on ZkNode. ZkMatrix can support transactions up to 2000 TPS and the entire process takes just 5 seconds. After the data is sent back to the main chain, to achieve maximum security, it is estimated a wait of 60 seconds for the data to be actually stored in the main chain.
| Go to ZkMatrix |
Get Started with Cortex
New to Cortex? Check out these community resources to get started with the AI on Blockchain Ecosystem of Cortex!
Whitepaper
Milestone / Roadmap
Ritchie 2022 Q1
Complete Solidity Compiler and MRT Upgrade.
Dijkstra 2022 Q2
Deploy the cross-chain bridge of the multi-node notary mechanism which is compatible with EVM. Publish papers for MRT quantification.
Bernoulli 2022 Q4
Complete the zk-Rollup for the transactions and contracts. Propose the cross-chain bridge of relay version.
Galileo 2023 Q4
Integrate the general distributed storage system architecture.
Roadmap
After the Mainnet launch, our roadmap is currently defined as follows.
Stage 0
Consistency: To upgrade and formalize MRT and CVM.
Stage 1
Scalability: Layer 1 or Layer 2 solutions to increase TPS (transactions per second) and DOPS (deterministic operations per second).
Stage 2
Privacy: Improve data and model privacy using crytographic solutions such as ZKP and TEE.