Hajime -- Decentralized AI Computing Power

Technical Framework

The Infinity Protocol architecture champions efficiency, speed, and sustainability, seamlessly integrating blockchain to empower AI startups with decentralized GPU access, ensuring swift, eco-friendly computing for Asia's innovator

Architecture Overview

Client Layer

This is the interactive face of the system, where various users, including the general public, AI startups (customers), and GPU providers (workers), interface with the platform. It is designed to be intuitive, offering a straightforward and engaging user experience.
‍GPU/CPU Devices: This includes all devices that provide processing power to the network.
‍PC/Mobile: Personal computing devices and mobile platforms used to access the network
‍Customized Device: Specialized hardware that may be integrated into the network for specific tasks.
‍Edge Devices: These are devices at the network's edge, enabling local data processing and sharing computational power. They can directly download and run ML models from the platform, reducing latency and allowing for real-time analytics.

Security Layer

Firewall: A system that enforces access control policies between networks, preventing unauthorized access to or from the network.
‍Rate Limiting: Mechanisms that control the amount of incoming requests a user can make to the API within a given timeframe, protecting against overuse or abuse.
API Gateway: The focal point for all incoming API requests, providing a single entry point for managing and routing API calls, as well as implementing authentication and authorization.

API Layer

Business API: Interfaces for business logic, user management, and other core functionalities.
‍Map API: Provides geolocation services to visually represent GPU providers on a map.
‍Device API: Manages the registration and status of devices providing computational resources.
Web3Auth API: Integrates blockchain-based authentication to verify user identities securely.

Service Layerr

Task Management Service: Manages the distribution and orchestration of computational tasks across the network.
Resource Allocation Service: Allocates available GPU/CPU resources to tasks based on requirements and availability.
Billing Service: Tracks resource usage and handles billing and payment processes.
Monitoring and Analytics Service: Provides insights into the performance of the network and its resources.
Messaging Queue Service: Manages communication between different services using a message queuing protocol.
Data Processing Service: Handles the computation and processing of data within the network.

Network Layer

Cluster VPN Mesh:
Connects the various nodes in the network securely, creating a mesh network that allows for efficient data transfer.
SD-WAN: Software-Defined Wide Area Network that provides optimal routing based on network conditions to ensure the best connection quality

Database Layer

Infrastructure Layer

Computing Pool: A collection of GPU/CPU resources pooled together from various devices available for computational tasks.
Kubernetes (K8S) Cluster: An orchestration system for managing containerized applications across a cluster of machines.
ML Task: Represents machine learning tasks that can be distributed across the computing pool to leverage the available computational resources.

Computing,
Power,
Democracy

Computing,
Power,
Democracy

Earn Money

Faster Computing

Lower Cost

Purer World

Technical Framework

Ready to get started?

Keen to connect?

Computing,Power,Democracy

Computing,Power,Democracy

Earn Money

Faster Computing

Lower Cost

Purer World

Technical Framework

Ready to get started?

Keen to connect?

Computing,
Power,
Democracy

Computing,
Power,
Democracy