Definitions of Key Terms & Concepts

This section explains the key terms and concepts essential for understanding the OmniTensor ecosystem, which blends decentralized AI and blockchain technology. The definitions provided are aimed at developers, AI researchers and blockchain enthusiasts, offering precise technical explanations along with practical applications within OmniTensor’s platform.

AI Compute Node

A node in the OmniTensor network that processes AI tasks like model inference and training. These nodes are run by users who contribute their hardware resources and are compensated with OMNIT tokens for completing AI-related tasks.

AI Grid as a Service (AI-GaaS)

A decentralized infrastructure model allowing users to access AI computing power, tools and services on demand. It taps into community-shared computational resources (GPUs/CPUs) to run AI models at scale, eliminating the need for centralized cloud providers. AI-GaaS supports seamless integration, training and deployment of AI applications.

AI Inference

This refers to using trained AI models to generate predictions or outputs from input data. On OmniTensor, inference is decentralized, leveraging distributed computing resources to reduce costs and improve accessibility.

AI Marketplace

A decentralized platform within OmniTensor where AI models and services can be bought, sold, or exchanged. Developers and researchers can publish their AI models for use or customization, earning OMNIT tokens based on model usage and adoption.

AI Model Hosting

A service within OmniTensor that hosts and delivers AI models across its network. Developers can publish models for either public or private use, monetizing them via OMNIT tokens.

Burn Mechanism

A deflationary system built into OmniTensor's blockchain that permanently removes OMNIT tokens from circulation, reducing the total supply to maintain long-term token value.

Data Layer

The part of OmniTensor that handles data collection, validation and usage for training AI models. The decentralized, community-driven data contributions are verified through blockchain to ensure quality and diversity for AI applications.

Decentralized Governance (DAO)

A governance model where OMNIT token holders vote on platform development and operational decisions. This decentralized structure ensures that the community has a say in the future direction of the OmniTensor platform.

Decentralized Physical Infrastructure Network (DePIN)

A blockchain-powered network allowing individuals to share unused computing resources (e.g., GPUs, CPUs) for AI tasks. DePIN enables scalable, cost-effective AI computation, rewarding contributors with OMNIT tokens.

dApp

Short for decentralized application, a software app that operates on a decentralized network like OmniTensor. Developers can create AI-driven dApps using OmniTensor’s infrastructure for tasks such as model deployment and data processing.

DualProof Consensus Mechanism

A hybrid consensus protocol combining Proof-of-Work (PoW) for AI compute tasks and Proof-of-Stake (PoS) for transaction validation. This dual system efficiently assigns compute-heavy tasks to PoW while PoS secures the network and validates transactions, ensuring both security and scalability.

EVM Chain (Ethereum Virtual Machine Chain)

OmniTensor's Layer 1 blockchain, built on Ethereum’s virtual machine (EVM). It manages smart contracts, token transactions and serves as the foundation for the OmniChain layer.

Governance

A decentralized decision-making framework in the OmniTensor ecosystem. It enables OMNIT token holders to propose and vote on platform updates and changes through a decentralized autonomous organization (DAO).

GPU Quantization

A technical process that optimizes GPU processing power for AI tasks. OmniTensor uses this technique to improve AI computations, enabling consumer-grade hardware to handle advanced workloads more efficiently.

Inference Network

A decentralized network within OmniTensor that facilitates AI inference tasks using preconfigured models. It helps lower costs and increase accessibility by distributing computation across community nodes.

Interoperability

The ability of OmniTensor’s infrastructure to operate smoothly across different blockchain systems (both Web2 and Web3). This capability is vital for integrating AI services into existing frameworks and allowing data and assets to move freely across platforms.

Model Fine-Tuning

The process of refining a pre-trained AI model with new or specific datasets. OmniTensor supports fine-tuning, enabling developers to enhance model accuracy and performance for specific tasks.

OMNIT Token

The native utility token of the OmniTensor ecosystem, used for transactions, incentivizing contributions (such as GPU sharing and data validation) and governing decentralized applications (dApps) on the platform. OMNIT tokens are also used for staking and participating in governance.

OmniChain

The Layer 2 blockchain within the OmniTensor platform, designed specifically for AI computation and cross-chain operations. It works alongside the Layer 1 EVM-based chain to offer scalable and secure AI services, handling AI model hosting, data validation and inference.

OmniPoints

A temporary reward system for participants contributing computing resources (GPUs/CPUs). These points can later be exchanged for OMNIT tokens during the Token Generation Event (TGE).

Proof-of-Stake (PoS)

Part of OmniTensor’s DualProof protocol, PoS is used on the Layer 1 EVM Chain for transaction validation and securing the network. Validators are rewarded for confirming AI computations performed by AI Compute Nodes.

Proof-of-Work (PoW)

A consensus mechanism used on the AI OmniChain to handle compute-intensive AI model tasks. PoW is a key element of the DualProof mechanism, ensuring accurate validation of decentralized AI computations.

SDK (Software Development Kit)

A set of tools and libraries offered by OmniTensor for developers to build, deploy and manage AI dApps. The SDK simplifies the development process by handling complexities related to AI model integration and blockchain interactions.

TGE (Token Generation Event)

The event marking the issuance and distribution of OMNIT tokens to participants who have earned OmniPoints. This event officially launches the tokenized economy within the OmniTensor ecosystem.

Tokenomics

The economic structure governing the issuance, distribution and utility of OMNIT tokens. Tokenomics includes incentives for participants, rewards for compute and data contributions and deflationary mechanisms like token burning to stabilize value.