Forward the Original Title ‘io.net研报：DePIN模式解决GPU资源困境，能否开启AI应用新时代？’

Introduction

In the digital age, computing power is crucial for technological advancement. It defines the ability of a computer to execute operations through its memory, processor speed, and the number of processors. This capability determines the efficiency of devices in executing tasks and their ability to handle multiple processes, thus affecting performance and cost. With the rise of artificial intelligence (AI) and deep learning models, there has been a surge in demand for accelerated computing power such as GPUs, leading to shortages worldwide.

The central processing unit (CPU), often considered the “brain” of the computer, along with the graphics processing unit (GPU), plays a vital role in determining the computing power of devices. A more powerful CPU can process more operations per second, while GPUs enhance CPU computing power by efficiently handling parallel tasks.

Source: Io.net

To address this growing demand, Io.net proposes an innovative solution: by leveraging the underutilized resources in data centers, cryptocurrency mining farms, and consumer GPUs, Io.net aims to build the world’s largest AI computing cloud. This strategy promises to provide nearly unlimited computing power at a fraction of the cost of traditional approaches, bridging the critical gap between rapidly growing demand for computing resources and limited supply. Io.net’s approach not only seeks to alleviate the current resource shortage issue but also supports the continued development of AI technology, ensuring the continuous advancement of digital capabilities.

Project overview

Io.net is committed to revolutionizing the cloud computing field through its decentralized GPU network designed specifically for machine learning. By establishing the world’s largest AI computing cloud, Io.net offers scalable, accessible, and cost-effective computing power, challenging the limitations of traditional cloud services. Through partnerships with entities like Filecoin and Render, as well as the upcoming IO token launch, Io.net enables AI/ML engineers to easily scale their projects within a vast GPU network.

The platform supports key AI development tasks such as batch inference, parallel training, and reinforcement learning, and has been optimized for Python workloads. With the introduction of the IO token, Io.net is moving towards a fully operational, cost-effective decentralized network. Leveraging DePIN technology, it can meet the diverse needs of GPU-dependent applications, providing a flexible, user-controlled computing environment. Io.net is setting new standards for decentralized cloud computing, leading the innovation wave in this field.

io.net features

IO Cloud

The IO Cloud aims to simplify the deployment and management of decentralized GPU clusters, providing machine learning engineers and developers with a scalable and flexible platform for accessing GPU resources without requiring significant hardware investment. The platform offers a user experience similar to traditional cloud services but with the added advantages of decentralized networks.

Highlights:

1. Scalability and Cost Efficiency: Committed to becoming the most cost-efficient GPU cloud platform, aiming to reduce AI/ML project costs by up to 90%.
2. Integration with IO SDK: Enhances the performance of AI projects through seamless integration, building a unified high-performance environment.
3. Global Coverage: Utilizes a CDN-like approach to distribute GPU resources globally, optimizing machine learning services and inference.
4. Support for RAY Framework: Utilizes the RAY distributed computing framework for scalable Python application development.
5. Exclusive Features: Provides private access to the OpenAI ChatGPT plugin, simplifying the deployment of training clusters.
6. Cryptomining Innovation: Aims to bring revolutionary innovation to the cryptomining industry by supporting the ML and AI ecosystem.

IO Worker

1. IO Worker is designed to optimize users’ supply operations in WebApps, including user account management, real-time activity monitoring, temperature and power tracking, installation support, wallet management, security assessment, and profitability analysis. It facilitates a more economically efficient and smooth AI learning process by bridging the gap between AI processing power demand and unused computing resource supply.

Highlights include:

1. Worker Dashboard: Provides a dashboard for real-time monitoring of connected devices, supporting functions such as device deletion and renaming.
2. Device Details Page: Displays comprehensive analytics of devices, including traffic, connection status, and job history.
3. Earnings and Rewards Page: Tracks earnings and job history, with transaction details available for query on SOLSCAN.
4. Add New Device Page: Simplifies the device connection process, supporting quick and easy integration of new devices.

IO Browser

1. IO Browser is designed as a comprehensive platform to provide users with in-depth insights into the operation of the io.net network, similar to how a blockchain explorer provides transparency for blockchain transactions. Its main goal is to enable users to monitor, analyze, and understand detailed information about the GPU cloud, ensuring complete visibility into network activity, statistics, and transactions while protecting the privacy of sensitive information.

Highlights include:

1. Browser Homepage: Provides insights into supply volume, verified suppliers, active hardware count, and real-time market pricing.
2. Clusters Page: Displays public information about deployed clusters in the network, along with real-time metrics and reservation details.
3. Device Page: Displays public details of devices connected to the network, providing real-time data and transaction tracking.
4. Real-time Clusters Monitoring: Provides instant insights into cluster status, health, and performance, ensuring users have the latest information.

Io.net’s architecture and technology

IO-SDK: Core Technology Engine

As a branch of Ray, the IO-SDK forms the cornerstone of IO.NET’s functionality. It supports task parallel execution and processing in multi-language environments. Its high compatibility with mainstream machine learning (ML) frameworks enables IO.NET to flexibly and efficiently meet diverse computational needs. This technological configuration, complemented by a finely defined technical system, ensures that the IO.NET platform can meet current demands while also possessing adaptability for future development.

Based on this, Io.net adopts a multi-layered architecture design, including:

User Interface Layer: Provides a visual frontend interface for users, including public websites, client areas, and GPU vendor areas, aiming to provide an intuitive and user-friendly experience.

Security Layer: Ensures the integrity and security of the system, integrating mechanisms such as network protection, user authentication, and activity logging.

API Layer: Serves as the communication hub for websites, vendors, and internal management, facilitating data exchange and the execution of various operations.

Backend Layer: Forms the core of the system, responsible for handling cluster/GPU management, client interaction, and automatic scaling operations.

Database Layer: Responsible for data storage and management, with the main storage handling structured data and caching used for temporary data processing.

Task Layer: Manages asynchronous communication and task execution to ensure the efficiency of data processing and circulation.

Infrastructure Layer: Forms the foundation of the system, including GPU resource pools, orchestration tools, and task execution/ML task processing, equipped with powerful monitoring solutions.

IO Tunneling Technology (IO Tunnels)

IO Tunnels establish secure connections from the client to remote servers, allowing engineers to bypass firewalls and NAT without complex configurations for remote access.

Workflow: IO workers first establish a connection with an intermediate server (i.e., the io.net server). Subsequently, the io.net server listens for connection requests from IO workers and engineer machines, facilitating data exchange through reverse tunneling technology.

Source: Io.net

Applications in io.net:

1. Through the io.net server, engineers can easily connect to IO workers without facing the challenges of network configuration, enabling remote access and management.

Advantages:

1. Convenience of Access: Direct connection to IO workers eliminates network barriers.
2. Security: Ensures communication security and protects data privacy.
3. Scalability and Flexibility: Efficiently manages multiple IO workers in different environments.

IO Network:

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

Characteristics of Mesh VPN Networks:

1. Decentralized Connectivity: Compared to traditional hub-and-spoke models, mesh VPN enables direct connections between nodes, enhancing redundancy, fault tolerance, and load distribution through multipath data transmission.
2. Advantages: Strong resistance to node failures, good scalability, reduced latency, and optimized traffic allocation.

Benefits to io.net:

1. Direct connections reduce communication latency, improving application performance.
2. No single point of failure ensures the network continues to operate even if a single node fails.
3. Enhances user privacy protection by increasing the difficulty of data tracking and analysis.
4. Easy addition of new nodes without affecting network performance.
5. Facilitates resource sharing and efficient processing between nodes.

Source: Io.net

Token Economics Overview

The $IO token is expected to be released in the second quarter of 2024 as the native cryptocurrency and protocol token of io.net, aiming to serve AI startups, developers, and GPU suppliers in the IO ecosystem, providing a range of functions to simplify payment processes, reward contributions, and facilitate efficient transactions of computational tasks.

For AI startups and developers:

1. Cluster Payments: $IO tokens simplify the payment process for cluster deployment and computational tasks.
2. Model Inference Gateway: Each model inference requires a microtransaction in $IO tokens.
3. IOSD Points: A utility token pegged to the US dollar, generated through the burning of $IO tokens, used for network transaction fees.

For suppliers:

1. GPU Rewards: Provides rewards for GPU contributions, including payment for rental time and incentives for hosting model inference on the network during idle time.

Io.net adopts a dual-token system of $IO and IOSD pegged to the US dollar to ensure stable transaction costs. The generation of IOSD is achieved through the burning of $IO tokens, with the process controlled by net emissions to maintain supply balance.

Tokenomics and Governance:

1. Burn-and-Mint Economics: Generates $IOSD through burning $IO, maintaining the monthly static supply of $IO, adjusted based on network usage.
2. Net Emission Mechanism: Adjusts the minting of $IO tokens based on the burning of IOSD, enabling the protocol to continue providing $IO rewards.
3. Governance Mechanism: $IO tokens serve as governance tokens, with holders influencing critical network decisions through staking, covering aspects such as fee settings and the introduction of new features.

Overview of Objectives and Design Principles:

1. io.net aims to connect global GPU resources, supporting computational tasks for artificial intelligence (AI) and machine learning (ML), driving network growth and balanced development through native asset incentive mechanisms.
2. Distribute rewards to GPU suppliers and provide loyalty points for AI engineers’ continued activity on the network.
3. Allow token holders to vote on key network decisions, advancing the implementation of decentralized governance structures.
4. Enable $IO token holders to stake investments to node operators or users, earning their reward shares.
5. A 5% fee will be levied during the network’s initial stage, with the proceeds used for $IO token burning activities or providing USD incentives, aimed at promoting network expansion and enhancing user participation.

Competitive product analysis

In the Web3 domain, numerous applications in the computing market are emerging, with prominent ones including Akash, Render, Nosana, Clore.ai, Inferx, Aioz, Fluence Labs, and GPU.NET. Despite fierce market competition, Io.net has quickly gained high attention in the industry, with an estimated valuation of $1 billion, thanks to the $30 million raised in its Series A funding round from prominent institutions such as Hack VC, Solana Labs, and cryptocurrency exchange platform OKX. In the field of DePIN, Io.net stands out with its leading cluster technology. This innovative feature allows Io.net to unite multiple computers into a powerful supercomputer, making it the only service capable of building decentralized clusters on a large scale, providing a crucial advantage for the development of complex AI solutions.

Source:Suyo on X

In 2023, Io.net received significant recognition at the Inventions Asia Awards, standing out among more than 110 projects from across Asia. Since early 2024, Io.net has demonstrated significant growth and engagement, with a waiting list of over 100,000 GPUs, providing over 14,000 hours of computing services. The network has registered over 20,000 GPUs and successfully built over 650 clusters.

It is worth noting that platforms like Io.net and Akash are dedicated to reforming traditional cloud computing in a decentralized manner, but they focus on different industry segments. While the Akash network provides a broad marketplace for cloud resources using container technology, Io.net focuses on providing high-performance GPU computing services for AI and machine learning needs, with unique cluster capabilities.

Additionally, the Render Network Foundation has pledged to allocate 300K RNDR from its total of 1.14M RNDR to new GPU node operators, with Io.net being the first partner in this initiative.

Bullish fundamentals

1. Io.net has established strong partnerships with Render Network and Solana, the latter being selected as the ideal choice for the DePIN project due to its high throughput and low fees, as reflected in the Messari 2024 report. It has also collaborated with Filecoin, opening up new revenue streams for decentralized storage providers while enhancing AI applications. Prior to this collaboration, Io.net had already received hardware donations from 1500 Filecoin members.
2. The team includes Angela Yi, a co-founder of a consulting firm for the Filecoin project and a Harvard alum, as well as Tory Green, the Chief Operating Officer with over 20 years of experience in venture capital and executive positions.
3. Io.net is poised to become a significant force in the DePIN field and is considered one of the most anticipated project launches of the first quarter. It is the only project capable of aggregating GPU resources from around the world into a unified distributed cluster, providing users with instant access services.
4. As the value of computing resources continues to increase, Io.net aims to reduce the cost of leasing GPU/CPU capacity, which is crucial for the scalability of AI and ML. The growth rate of computing resource demand exceeds the growth rate of supply, indicating an increasingly widening gap between demand and supply.
5. It is expected that by 2030, particularly the demand for computing resources required for model training, will become a huge market, potentially accounting for 1% of the US GDP.

Bearish fundamentals

1. The decentralized computing market is becoming increasingly competitive, with mature projects like Akash Network and emerging projects continuously entering the market. Intense competition may affect Io.net’s ability to retain users.
2. There are significant technical challenges in the large-scale creation and management of decentralized clusters.
3. Despite growing interest in decentralized computing solutions, widespread adoption is still in its early stages. The slower-than-expected adoption rates among AI startups, developers, and GPU suppliers may limit growth.
4. Legal changes in cryptocurrency and decentralized networks, as well as the use of the $IOSD stablecoin, may attract regulatory attention.
5. The dual-token system involving $IO and $IOSD may be too complex for users.

Conclusion

io.net is poised to shine in the promising AI computing market, backed not only by a skilled technical team but also by prominent entities such as Multicoin Capital, Solana Ventures, OKX Ventures, Aptos Labs, Delphi Digital, and others. Given the current situation and its upcoming token issuance, we believe this is a project worth closely monitoring, but we always remind investors to proceed with caution.

Disclaimer：

1. This article is reprinted from [Panewslab]. Forward the Original Title‘io.net研报：DePIN模式解决GPU资源困境，能否开启AI应用新时代？’.All copyrights belong to the original author [0xGreythorn]. 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.