1. Project Overview

io.net is a distributed GPU system based on Solana, Render, Ray, and Filecoin, aiming to address the computational challenges in AI and machine learning by utilizing decentralized GPU resources.

By aggregating underutilized computing resources from independent data centres, cryptocurrency miners, and surplus GPUs from projects like Filecoin and Render, io.net tackles the issue of insufficient computing power. This enables engineers to access a large amount of computing power in a system that is easily accessible, customizable, and cost-effective. Additionally, io.net introduces a distributed physical infrastructure network (DePIN), combining resources from various providers. This approach allows engineers to acquire significant computing power in a customizable, cost-effective, and easy-to-implement manner. io. cloud currently boasts over 95,000 GPUs and more than 1,000 CPUs, supporting quick deployment, hardware selection, geographic location, and providing a transparent payment process.

2. Core Mechanisms

2.1 Decentralized Resource Aggregation

One of io.net’s core functions is its decentralized resource aggregation, enabling the platform to leverage distributed GPU resources globally to support AI and machine learning tasks. This strategy aims to optimize resource usage, reduce costs, and enhance accessibility.

Here’s a detailed breakdown:

2.1.1 Benefits

• Cost Efficiency: By utilizing underused GPU resources, io.net offers computing power at lower costs than traditional cloud services, which is crucial for data-intensive AI applications that typically require vast amounts of computational power.
• Scalability and Flexibility: The decentralized model allows io.net to expand its resource pool easily without relying on a single vendor or data centre, offering users the flexibility to choose resources that best meet their needs.

2.1.2 How It Works

• Diverse Resource Sources: io.net aggregates GPU resources from various sources, including independent data centres, individual cryptocurrency miners, and surplus resources from projects like Filecoin and Render.
• Technological Implementation: The platform uses blockchain technology to track and manage these resources, ensuring transparent and fair resource allocation. Blockchain also automates payments and incentives for users who contribute additional computing power to the network.

2.1.3 Steps Involved

• Resource Discovery and Registration: Resource providers (e.g., GPU owners) register their devices on the io.net platform. The platform verifies the performance and reliability of these resources to ensure they meet specific standards and requirements.
• Resource Pooling: Verified resources are added to a global pool available for rent by platform users. Smart contracts automatically manage the distribution and management of resources, ensuring transparency and efficiency.
• Dynamic Resource Allocation: When users initiate a computational task, the platform dynamically allocates resources based on the task’s requirements (e.g., computing power, memory, network bandwidth). Resource allocation considers cost efficiency and geographic location to optimize task execution speed and cost.

2.2 Dual Token Economic System

io.net’s dual token economic system is a key feature designed to incentivize network participants and ensure the platform’s efficiency and sustainability. The system includes two tokens: $IO and $IOSD, each with distinct roles. Here’s a detailed overview:

2.2.1 $IO Token

$IO is the primary functional token of the io.net platform, used for various network transactions and operations. Its main uses include:

• Payments and Fees: Users pay for computing resource rentals, including GPU usage fees, with $IO. It is also used for various services and fees on the network.
• Resource Incentives: $IO tokens are awarded to those who provide GPU computing power or participate in maintaining the network, encouraging continuous resource contribution.
• Governance: $IO token holders can participate in the governance decisions of the io.net platform, influencing the platform’s future development and policy adjustments through voting rights.

2.2.2 $IOSD Token

$IOSD is a stablecoin pegged to the US dollar, designed to provide a stable value storage and transaction medium on the io.net platform. Its main functions include:

• Value Stability: Pegged to the US dollar at a 1:1 ratio, $IOSD offers users a payment method that avoids cryptocurrency market volatility.
• Transaction Convenience: Users can pay platform fees, such as computing resource fees, with $IOSD, ensuring stability and predictability in transactions.
• Fee Coverage: Certain network operations or transaction fees can be paid with $IOSD, simplifying the fee settlement process.

2.2.3 Interaction of the Dual Token System

io.net’s dual token system supports network operations and growth through several interactions:

• Resource Provider Incentives: Resource providers (e.g., GPU owners) earn $IO tokens for contributing their devices to the network. These tokens can be used to purchase computing resources or traded in the market.
• Fee Payments: Users pay for computing resource usage with $IO or $IOSD. Using $IOSD avoids the risks associated with cryptocurrency volatility.
• Economic Activity Incentives: The circulation and use of $IO and $IOSD stimulate economic activity on the io.net platform, increasing network liquidity and participation.
• Governance Participation: $IO tokens also serve as governance tokens, allowing holders to participate in the platform’s governance, such as proposing and voting on decisions.

2.3 Dynamic Resource Allocation and Scheduling

io.net’s dynamic resource allocation and scheduling are crucial for efficiently managing and optimizing the use of computing resources to meet users’ diverse computational needs. This system ensures that computational tasks are executed on the most suitable resources in an intelligent and automated manner, maximizing resource utilization and performance.

Here’s a detailed look at this mechanism:

2.3.1 Dynamic Resource Allocation Mechanism

Resource Identification and Classification:

• When resource providers connect their GPUs or other computing resources to the io.net platform, the system identifies and classifies these resources by assessing performance indicators like processing speed, memory capacity, and network bandwidth.
• These resources are then tagged and archived for dynamic allocation based on different task requirements.

Demand Matching:

• Users submit computational tasks to io.net, specifying requirements like required computing power, memory size, and budget constraints.
• The platform’s scheduling system analyzes these requirements and selects matching resources from the pool.

Intelligent Scheduling Algorithm:

• Advanced algorithms automatically match the most suitable resources with the submitted tasks, considering resource performance, cost efficiency, geographic location (to reduce latency), and user preferences.
• The scheduling system monitors resources’ real-time status, like availability and load, to dynamically adjust resource allocation.

2.3.2 Scheduling and Execution

Task Queuing and Priority Management:

• All tasks are queued based on priority and submission time. The system handles the task queue using preset or dynamically adjusted priority rules.
• Urgent or high-priority tasks receive quick responses, while long-term or cost-sensitive tasks may be executed during low-cost periods.

Fault Tolerance and Load Balancing:

• The dynamic resource allocation system includes fault tolerance mechanisms, ensuring tasks can migrate to other healthy resources for continued execution even if some resources fail.
• Load balancing techniques ensure no single resource is overloaded, optimizing network performance through reasonable task load distribution.

Monitoring and Adjustment:

• The system continuously monitors task execution status and resource conditions, analyzing key performance indicators like task progress and resource consumption in real time.
• Based on this data, the system may automatically readjust resource allocation to optimize task execution efficiency and resource utilization.

2.3.3 User Interaction and Feedback

• Transparent User Interface: io.net provides an intuitive user interface where users can easily submit tasks, view task status, and adjust requirements or priorities.
• Feedback Mechanism: Users can provide feedback on task execution results, and the system adjusts future task resource allocation strategies based on feedback to better meet user needs.

3. System Architecture

3.1 IO Cloud

IO Cloud simplifies deploying and managing decentralized GPU clusters, offering scalable and flexible GPU resources for machine learning engineers and developers without significant hardware investment. This platform delivers an experience similar to traditional cloud services but with decentralized network benefits. Highlights include:

• Scalability and Cost-effectiveness: Targets a cost-effective GPU cloud, potentially reducing AI/ML project costs by up to 90%.
• Integration with IO SDK: Enhances AI project performance through seamless integration, creating a unified high-performance environment.
• Global Coverage: Utilizes distributed GPU resources to optimize machine learning services and inference, similar to a CDN.
• RAY Framework Support: Supports scalable Python application development using the RAY distributed computing framework.
• Exclusive Features: Provides private access to the OpenAI ChatGPT plugin, facilitating the deployment of training clusters.
• Cryptocurrency Mining Innovation: Aims to innovate cryptocurrency mining by supporting the machine learning and AI ecosystem.

3.2 IO Worker

IO Worker aims to simplify and optimize provisioning operations for WebApp users, including user account management, real-time activity monitoring, temperature and power consumption tracking, installation support, wallet management, security, and profitability analysis. Highlights:

• Worker Homepage: Offers a dashboard for real-time monitoring of connected devices, with options to remove and rename devices.
• Device Details Page: Provides comprehensive device analysis, including traffic, connection status, and work history.
• Earnings and Rewards Page: Tracks earnings and work history, with transaction details accessible on SOLSCAN.
• Add New Device Page: Simplifies the device connection process, supporting quick and easy integration.

3.3 IO Explorer

IO Explorer provides users with in-depth insights into io.net network operations, similar to blockchain explorers for blockchain transactions. It aims to enable users to monitor, analyze, and understand detailed information about the GPU cloud, ensuring visibility into network activities, statistics, and transactions while protecting sensitive information. Advantages:

• Explorer Homepage: Offers insights into supply, verified suppliers, active hardware, and real-time market pricing.
• Cluster Page: Displays public information about deployed clusters in the network, along with real-time metrics and booking details.
• Device Page: Shows public details of devices connected to the network, providing real-time data and transaction tracking.
• Real-time Cluster Monitoring: Provides instant insights into cluster status, health, and performance, ensuring users receive the latest information.

3.4 IO-SDK

IO-SDK, derived from a branch of Ray technology, is the foundational technology of io.net. It enables parallel task execution and multi-language processing and is compatible with major machine learning frameworks. This setup ensures that IO.NET can meet current demands and adapt to future changes.

The multi-layer architecture includes:

• User Interface: The visual frontend for users, including the public website, customer area, and GPU provider area. Designed to be intuitive and user-friendly.
• Security Layer: Ensures system integrity and security, including network protection, user authentication, and activity logging.
• API Layer: Acts as the communication hub for the website, providers, and internal management, facilitating data exchange and operations.
• Backend Layer: The system’s core, handling operations such as cluster/GPU management, customer interactions, and auto-scaling.
• Database Layer: Stores and manages data, with primary storage for structured data and caching for temporary data.
• Task Layer: Manages asynchronous communication and tasks, ensuring efficiency in execution and data flow.
• Infrastructure Layer: The foundation, containing GPU pools, orchestration tools, and execution/ML tasks, is equipped with robust monitoring solutions.

3.5 IO Tunnels

• IO Tunnels utilize reverse tunnelling technology to create secure connections from the client to remote servers, allowing engineers to bypass firewalls and NAT for remote access without complex configurations.
• Workflow: The IO Worker connects to the intermediate server (io.net server). The io.net server then listens for connections from IO Worker and engineer machines, facilitating data exchange through reverse tunnelling.

Application in io.net

• Engineers connect to IO Workers via the io.net server, simplifying remote access and management without network configuration challenges.
• Advantages: Convenient Access: Direct access to IO Workers, eliminating network barriers.
• Security: Ensures protected communication, and maintains data privacy.
• Scalability and Flexibility: Effectively manages multiple IO Workers in different environments.

3.6 IO Network

• IO Network adopts a mesh VPN architecture to provide ultra-low latency communication between antMiner nodes.
  

Mesh VPN Network:

• Decentralized Connectivity: Unlike traditional star models, a mesh VPN directly connects nodes, offering enhanced redundancy, fault tolerance, and load distribution.
• Advantages: Strong resistance to node failures, high scalability, low latency, and optimized traffic distribution.

Benefits of io.net:

• Direct connections reduce latency, optimizing application performance.
• No single point of failure ensures network operation even if individual nodes fail.
• Enhances user privacy by making data tracking and analysis more challenging.
• Adding new nodes does not affect performance.
• Resource sharing and processing are more efficient between nodes.

4. $IO Token

4.1 Basic Framework of $IO Token

• Fixed Supply:

The total supply of $IO tokens is capped at 800 million, ensuring stability and preventing inflation.

• Distribution and Incentives:
• Initially, 300 million $IO tokens will be distributed. The remaining 500 million will be awarded to suppliers and their stakeholders over 20 years.
• Rewards are released hourly, following a diminishing model (starting at 8% in the first year, decreasing by 1.02% monthly, roughly 12% annually) until the 800 million cap is reached.
• Burn Mechanism:

$IO has a programmed token burn system where io.net uses revenue from the IOG network to buy and burn $IO tokens. The burn quantity adjusts based on $IO’s price, creating deflationary pressure.

4.2 Fees and Earnings

• Usage Fees:

io.net charges users and suppliers various fees, including booking and payment fees for computing power. These fees support the network’s financial health and $IO’s market circulation.

• Payment Fees:

A 2% fee applies to USDC payments; no fee for $IO payments.

• Supplier Fees:

Suppliers also pay booking and payment fees when receiving payments, similar to users.

4.3 Ecosystem

• GPU Renters (Users):

Machine learning engineers seeking GPU computing power on the IOG network use $IO to deploy GPU clusters, cloud gaming instances, and build applications like Unreal Engine 5 pixel streaming. Users also include individuals performing serverless model inference on BC8.ai and future applications hosted by io.net.

• GPU Owners (Suppliers):

Independent data centers, crypto mining farms, and professional miners offering underutilized GPU computing power on the IOG network.

• IO Token Holders (Community):

The community provides crypto-economic security and incentives to coordinate mutually beneficial actions, fostering network growth and adoption.

4.4 Specific Allocation

• Community: 50% for rewarding community members and encouraging platform participation and growth.
• R&D Ecosystem: 16% for supporting R&D and ecosystem building, including partners and third-party developers.
• Initial Core Contributors: 11.3% for rewarding early-stage contributors.
• Early Backers: Seed: 12.5% for early seed investors, rewarding their early support.
• Early Backers: Series A: 10.2% for Series A investors, rewarding their contributions in the early development stages.

4.5 Halving Mechanism

• 2024 to 2025: 6,000,000 $IO tokens released annually.
• 2026 to 2027: Annual release halved to 3,000,000 $IO tokens.
• 2028 to 2029: Annual release halved again to 1,500,000 $IO tokens.

5. Team/Partnerships/Funding

io.net’s leadership team brings diverse skills and experience. Tory Green, the COO, was previously COO of Hum Capital and Director of Corporate Development and Strategy at Fox Mobile Group. Ahmad Shadid, the Founder and CEO, was a Quantitative Systems Engineer at WhalesTrader. Garrison Yang, the Chief Strategy Officer and CMO was VP of Growth and Strategy at Ava Labs, with a degree in Environmental Health Engineering from UC Santa Barbara.

In March, io.net raised $30 million in Series A funding, led by Hack VC, with participation from Multicoin Capital, 6th Man Ventures, M13, Delphi Digital, Solana Labs, Aptos Labs, Foresight Ventures, Longhash, SevenX, ArkStream, Animoca Brands, Continue Capital, MH Ventures, and OKX. Industry leaders such as Solana founder Anatoly Yakovenko, Aptos founders Mo Shaikh and Avery Ching, Animoca Brands’ Yat Siu, and Perlone Capital’s Jin Kang also invested.

6. Project Evaluation

6.1 Market Analysis

io.net is a decentralized computing network built on the Solana blockchain, focusing on integrating underutilized GPU resources to provide powerful computing capabilities. This project operates mainly in the following areas:

• Decentralized Computing:

io.net has developed a decentralized physical infrastructure network (DePIN) that leverages GPU resources from various sources (such as independent data centers and cryptocurrency miners). This decentralized approach aims to optimize computing resource utilization, reduce costs, and enhance accessibility and flexibility.

• Cloud Computing:

Although io.net uses a decentralized approach, it offers services similar to traditional cloud computing, such as GPU cluster management and scaling for machine learning tasks. io.net aims to deliver an experience similar to traditional cloud services but with the efficiency and cost advantages of a decentralized network.

• Blockchain Applications:

As a blockchain-based project, io.net uses blockchain features like security and transparency to manage resources and transactions within the network.

Similar projects in terms of functionality and goals include:

• Golem: A decentralized computing network where users can rent or lease unused computing resources. Golem aims to create a global supercomputer.
• Render: Uses a decentralized network to provide graphic rendering services, leveraging blockchain technology to enable content creators to access more GPU resources, speeding up the rendering process.
• iExec RLC: Creates a decentralized marketplace allowing users to rent their computing resources, supporting various applications through blockchain technology, including data-intensive applications and machine learning workloads.

6.2 Project Advantages

• Scalability: io.net is designed as a highly scalable platform to meet customers’ bandwidth needs, enabling teams to scale workloads on the GPU network easily without significant adjustments.
• Batch Inference and Model Serving: The platform supports parallel inference on data batches, allowing machine learning teams to deploy workflows on a distributed GPU network.
• Parallel Training: To overcome memory limitations and sequential workflows, io.net utilizes a distributed computing library to parallelize training tasks across multiple devices.
• Parallel Hyperparameter Tuning: io.net optimizes the scheduling and search patterns by leveraging the inherent parallelism of hyperparameter tuning experiments.
• Reinforcement Learning (RL): Using open-source RL libraries, io.net supports highly distributed RL workloads and offers a simple API.
• Instant Accessibility: Unlike traditional cloud services with long deployment times, io.net Cloud provides instant access to GPU supply, enabling users to launch projects within seconds.
• Cost Efficiency: io.net is designed as an affordable platform suitable for various user categories. Currently, the platform is approximately 90% more cost-efficient than competing services, providing significant savings for machine learning projects.
• High Security and Reliability: The platform promises top-tier security, reliability, and technical support, ensuring a secure and stable environment for machine learning tasks.
• Ease of Implementation: io.net Cloud eliminates the complexity of building and managing infrastructure, allowing any developer or organization to seamlessly develop and scale AI applications.

6.3 Project Challenges

• Technical Complexity and User Adoption:
• Challenge: While decentralized computing offers significant cost and efficiency advantages, its technical complexity may pose a considerable barrier for non-technical users. Users need to understand how to operate a distributed network and effectively utilize distributed resources.
• Impact: This could limit the platform’s widespread adoption, particularly among users less familiar with blockchain and decentralized computing.
• Network Security and Data Privacy:
• Challenge: Despite the enhanced security and transparency provided by blockchain, the openness of decentralized networks may make them more susceptible to cyberattacks and data breaches.
• Impact: This requires io.net to continually strengthen its security measures to ensure the confidentiality and integrity of user data and computing tasks, which is crucial for maintaining user trust and platform reputation.
• Performance and Reliability:
• Challenge: While io.net aims to provide efficient computing services through decentralized resources, coordinating across different geographical locations and varying hardware quality can present performance and reliability challenges.
• Impact: Any performance issues due to hardware mismatches or network latency could affect customer satisfaction and the platform’s overall effectiveness.
• Scalability of Operations:
• Challenge: Although io.net is designed as a highly scalable network, effectively managing and scaling decentralized resources globally remains a significant technical challenge in practice.
• Impact: Continuous technical innovation and management improvements are needed to maintain network stability and responsiveness amid rapidly growing user and computing demands.
• Competition and Market Acceptance:
• Challenge: io.net faces competition in the blockchain and decentralized computing market. Other platforms like Golem, Render, and iExec offer similar services, and the market’s rapid evolution could quickly alter the competitive landscape.
• Impact: To stay competitive, io.net needs continuous innovation and improvement in its services’ uniqueness and value to attract and retain users.

1. Conclusion

io.net sets a new standard in the modern cloud computing field with its innovative decentralized computing network and blockchain-based architecture. By aggregating underutilized GPU resources worldwide, io.net provides unprecedented computing power, flexibility, and cost efficiency for machine learning and AI applications. The platform not only makes large-scale machine learning project deployment more accessible and economical but also offers robust security and scalable solutions for various users. Despite challenges such as technical complexity, network security, performance stability, and market competition, if io.net can overcome these hurdles and cultivate a vibrant ecosystem, it has the potential to fundamentally reshape how we access and utilize computing power in the Web3 era. However, like any emerging technology, its long-term success will depend on continuous development, adoption, and its ability to navigate the evolving landscape of blockchain-based infrastructure.

Disclaimer：

1. This article is reprinted from[链茶馆]. All copyrights belong to the original author [茶馆小二儿]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
3. Translations of the article into other languages are done by the Gate Learn team. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.