OmniTensor
  • Welcome
  • Introduction
    • What is OmniTensor?
    • Vision & Mission
    • Key Features
    • Why OmniTensor?
      • Current Challenges in AI
      • How OmniTensor Addresses These Challenges
  • Core Concepts
    • AI Grid as a Service (AI-GaaS)
      • Overview of AI-GaaS
      • Benefits of AI-GaaS
      • Use Cases of AI-GaaS
    • Decentralized Physical Infrastructure Network (DePIN)
      • GPU Sharing Model
      • Incentive Mechanisms
    • AI OmniChain
      • Layer 1 and Layer 2 Integration
      • AI Model Marketplace and Interoperability
    • DualProof Consensus Mechanism
      • Proof-of-Work (PoW) for AI Compute
      • Proof-of-Stake (PoS) for Validation
    • OMNIT Token
      • Overview
      • Utility
      • Governance
  • Tokenomics
    • Token Allocations
    • Token Locks
    • ERC20 Token
    • Audit
  • OmniTensor Infrastructure
    • L1 EVM Chain
      • Overview & Benefits
      • Development Tools & API
    • AI OmniChain
      • Interoperability
      • Scalability
      • Decentralized Data & Model Management
    • Nodes & Network Management
      • AI Consensus Validator Nodes
      • AI Compute Nodes (GPUs)
  • Roadmap & Updates
    • Roadmap
    • Future Features
  • PRODUCTS
    • AI Model Marketplace
    • dApp Store
    • Data Layer
    • Customizable Solutions
    • AI Inference Network
  • For the Community
    • Contributing to OmniTensor
      • Sharing Your GPU
      • Data Collection & Validation
    • Earning OMNIT Tokens
      • Computation Rewards
      • Data Processing & Validation Rewards
    • Community Incentives & Gamification
      • Participation Rewards
      • Leaderboards & Competitions
  • For Developers
    • Building on OmniTensor
      • dApp Development Overview
      • Using Pre-trained AI Models
    • SDK & Tools
      • OmniTensor SDK Overview
      • API Documentation
    • AI Model Training & Deployment
      • Training Custom Models
      • Deploying Models on OmniTensor
    • Decentralized Inference Network
      • Running AI Inference
      • Managing and Scaling Inference Tasks
    • Advanced Topics
      • Cross-Chain Interoperability
      • Custom AI Model Fine-Tuning
  • For Businesses
    • AI Solutions for Businesses
      • Ready-Made AI dApps
      • Custom AI Solution Development
    • Integrating OmniTensor with Existing Systems
      • Web2 & Web3 Integration
      • API Usage & Examples
    • Privacy & Security
      • Data Encryption & Privacy Measures
      • Secure AI Model Hosting
  • Getting Started
    • Setting Up Your Account
    • Installing SDK & CLI Tools
  • Tutorials & Examples
    • Building AI dApps Step by Step
    • Integrating AI Models with OmniTensor
    • Case Studies
      • AI dApp Implementations
      • Real-World Applications
  • FAQ
    • Common Questions & Issues
    • Troubleshooting
  • Glossary
    • Definitions of Key Terms & Concepts
  • Community and Support
    • Official Links
    • Community Channels
  • Legal
    • Terms of Service
    • Privacy Policy
    • Licensing Information
Powered by GitBook
On this page
  1. Core Concepts
  2. DualProof Consensus Mechanism

Proof-of-Work (PoW) for AI Compute

OmniTensor’s framework incorporates a Proof-of-Work (PoW) element to manage AI computing tasks. When an AI Compute Node, which represents decentralized hardware (such as community-contributed GPUs), is assigned an AI task, it carries out the necessary computations. These tasks can range from AI inference to model training, similar to the cryptographic processes used in traditional blockchain systems. However, instead of solving puzzles, the PoW mechanism is applied to complete valuable AI computations. This approach allows the system to tap into the processing power of a distributed GPU network to handle complex AI operations.

The PoW mechanism converts each task into proof of the work done on AI computations, forming a decentralized AI network. After a computation is finished, it must be verified to confirm the accuracy of the outcome, which brings in the Proof-of-Stake (PoS) component of the framework.

PreviousDualProof Consensus MechanismNextProof-of-Stake (PoS) for Validation

Last updated 7 months ago