Forward the Original Title’MIIX Capital: io.net项目研究报告’

1. Project Status

1.1 Operational Overview

io.net is a decentralized GPU network aimed at providing computing power for machine learning (ML). It assembles computing resources from over 1 million GPUs sourced from independent data centers, cryptocurrency miners, and projects like Filecoin and Render.

Its goal is to combine these 1 million GPUs into a DePIN (Decentralized Physical Infrastructure Network) to create an enterprise-grade, decentralized distributed computing network. By pooling global idle computing resources (mainly GPUs), it offers AI engineers more affordable, accessible, and flexible network computing services.

For users, it functions like a decentralized marketplace for idle global GPU resources, allowing AI engineers or teams to customize and purchase the GPU computing services they need based on their requirements.

1.2 Team Background

Ahmad Shadid is the founder and CEO, previously a quantitative systems engineer at WhalesTrader.

Garrison Yang is the Chief Strategy Officer and Chief Marketing Officer, previously Vice President of Growth and Strategy at Ava Labs.

Tory Green is the Chief Operating Officer, formerly the COO of Hum Capital and Director of Corporate Development and Strategy at Fox Mobile Group.

Angela Yi is the Vice President of Business Development, a Harvard University graduate, responsible for planning and executing key strategies in sales, partnerships, and vendor management.

In 2020, when Ahmad Shadid was building a GPU computing network for the machine learning quantitative trading company Dark Tick, the trading strategies were close to high-frequency trading, requiring a vast amount of computing power. The high cost of GPU services from cloud providers became a significant challenge for them. The immense demand for computing power and the high costs prompted them to pursue decentralized distributed computing resources. They later gained attention at the Austin Solana Hacker House. Thus, io.net emerged from the team’s own challenges, offering a solution and expanding their business.

1.3 Products/Technology

Market users face several challenges:

Limited availability Accessing hardware through cloud services like AWS, GCP, or Azure often takes weeks, and popular GPU models are frequently unavailable.

Limited choices: Users have minimal flexibility in terms of GPU hardware, location, security level, latency, etc.

High costs: Acquiring high-quality GPUs is costly, with monthly expenses reaching hundreds of thousands of dollars for training and inference.

Solution:

io.net addresses these challenges by aggregating underutilized GPU resources (such as independent data centers, cryptocurrency miners, and projects like Filecoin and Render) into DePIN. This integration allows engineers to access substantial computing power within the system. It enables ML teams to build inference and model service workflows across distributed GPU networks, utilizing distributed computing libraries to orchestrate and batch-train jobs for parallelization across multiple distributed devices.

Additionally, io.net utilizes distributed computing libraries with advanced hyperparameter tuning to examine optimal results, optimize scheduling, and specify search patterns easily. It also employs an open-source reinforcement learning library supporting production-grade, highly distributed RL workloads and a simple API.

Product Components:

IO Cloud: Designed to deploy and manage decentralized GPU clusters on-demand, seamlessly integrating with IO-SDK to offer a comprehensive solution for scaling AI and Python applications. It provides unlimited computing power while simplifying the deployment and management of GPU/CPU resources.

IO Worker: Offers users a comprehensive and user-friendly interface to efficiently manage their GPU node operations through an intuitive web application. Features include account management, monitoring of computing activities, real-time data display, temperature and power tracking, installation assistance, wallet management, security measures, and profitability calculations.

IO Explorer: Primarily provides users with comprehensive statistical data and visualizations of various aspects of the GPU cloud, enabling users to easily monitor, analyze, and understand the complex details of the io.net network. It offers comprehensive visibility into network activity, key statistics, data points, and reward transactions.

Product Features:

Decentralized computing network: io.net adopts a decentralized computing model, distributing computing resources globally to enhance efficiency and stability.

Low-cost access: io.net Cloud offers lower access costs compared to traditional centralized services, enabling more ML engineers and researchers to access computing resources.

Distributed cloud clusters: The platform provides a distributed cloud cluster where users can select suitable computing resources according to their needs and distribute tasks to different nodes for processing.

Support for ML tasks: io.net Cloud focuses on providing computing resources for ML engineers, facilitating easier model training, data processing, and other tasks.

1.4 Development Roadmap

https://developers.io.net/docs/product-timeline

According to the information disclosed in the io.net whitepaper, the project’s product roadmap is as follows: January-April 2024: Full release of V1.0, focusing on decentralizing the io.net ecosystem to enable self-hosting and self-replication.

1.5 Funding Information

According to public news sources, on March 5, 2024, io.net announced the completion of a $30 million Series A financing round. Hack VC led the round, 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, Sandbox Games, and others. [1] It’s worth noting that following this financing round, io.net reached an overall valuation of $1 billion.

2. Market Data

2.1 Official Website

According to the official website data from January 2024 to March 2024, the total number of visits is 5.212 million, with an average monthly visit of 1.737 million and a bounce rate of 18.61% (relatively low). User visits are evenly distributed across various regions, and direct visits and search visits account for over 80%. This may indicate that the proportion of dirty data in the visitor data is low. Users have a basic understanding of io.net and are willing to further explore and interact on the website.

2.2 Social Media Communities

3. Competitive Analysis

3.1 Competitive Landscape

io.net’s core business is closely related to decentralized AI computing power, with its major competitors being traditional cloud service providers represented by AWS, Google Cloud, and Microsoft Azure. According to the “2022-2023 Global Computing Power Index Assessment Report” jointly compiled by IDC, CCID Consulting, and the Global Industry Research Institute of Tsinghua University, the global artificial intelligence computing market is expected to grow from $19.5 billion in 2022 to $34.66 billion in 2026. [2]

Comparing the sales revenue of leading global cloud computing vendors: In 2023, AWS had cloud service sales revenue of $90.8 billion, Google Cloud had a sales revenue of $33.7 billion, and Microsoft Azure had a sales revenue of $96.8 billion. [3] These three companies together hold around 66% of the global market share. Additionally, the market value of these three giant companies is all above one trillion dollars.

https://www.alluxio.io/blog/maximize-gpu-utilization-for-model-training/

In stark contrast to the high revenue of cloud service providers, improving GPU utilization has become a focal issue. According to a survey on AI infrastructure, most GPU resources are underutilized—around 53% of respondents believe that 51-70% of GPU resources are underutilized, 25% estimate utilization rates at 85%, and only 7% believe utilization exceeds 85%. For io.net, the significant demand for cloud computing and the issue of underutilized GPU resources present market opportunities.

3.2 Advantage Analysis

https://twitter.com/eli5_defi/status/1768261383576289429

io.net’s greatest competitive advantage lies in its ecological positioning or first-mover advantage. According to the data provided by the official source, io.net currently possesses a GPU cluster totaling more than 40,000, CPU total exceeding 5,600, and over 69,000 Worker Nodes. The deployment time for 10,000 GPUs is less than 90 seconds, and its prices are 90% cheaper compared to competitors, with a valuation of $1 billion.

io.net not only offers customers low-cost and instant online services compared to centralized cloud service providers at 1-2% of the cost but also provides additional launch incentives for computing power providers through the upcoming IO token, facilitating the goal of connecting 1 million GPUs.

Furthermore, compared to other DePIN computing projects, io.net focuses on GPU computing power, with its GPU network scale surpassing similar projects by over 100 times. io.net is also the first project in the blockchain industry to integrate state-of-the-art ML technology stacks (such as Ray clusters, Kubernetes clusters, and giant clusters) into GPU DePIN projects and put them into large-scale practice, placing it in a leading position not only in terms of GPU quantity but also in technology application and model training capabilities.

With the continuous development of io.net, if it can increase GPU capacity to compete with centralized cloud service providers at 500,000 concurrent GPUs, it can provide services similar to Web 2 at lower costs. There is also an opportunity to gradually establish its core position as a decentralized GPU network leader and settlement layer in collaboration with major DePIN and AI players (including Render Network, Filecoin, Solana, Ritual, etc.), bringing vitality to the entire Web 3xAI ecosystem.

3.3 Risks and Issues

io.net is an emerging computing resource integration and distribution platform deeply integrated with Web3. Its business overlaps significantly with traditional cloud service providers, posing risks and obstacles both in terms of technology and market positioning.

Technical Security Risks: As a nascent platform, io.net has not undergone large-scale application testing and lacks demonstrated capabilities to prevent and respond to malicious attacks. With the massive influx, distribution, and management of computational resources, there’s a lack of corresponding experience or practical validation, making it susceptible to common technical issues such as compatibility, robustness, and security. Any issues that arise could be potentially fatal for io.net, as customers prioritize their security and stability and are unwilling to bear the consequences.

Slow Market Expansion: io.net competes directly with traditional cloud service providers such as AWS, Google Cloud, and Alicloud, as well as second-tier or third-tier service providers. Despite its cost advantages, io.net’s service and market systems targeting B-class customers are still in their infancy, presenting a significant difference from the existing market operations in the Web3 industry. Therefore, its progress in market expansion may not be ideal, directly impacting its project valuation and token market performance.

Latest Security Incident

On April 25th, io.net’s founder and CEO Ahmad Shadid tweeted about a security incident involving io.net’s metadata API. Attackers exploited the mapping from user IDs to device IDs, leading to unauthorized updates of metadata. While this vulnerability did not affect GPU access, it did impact the metadata displayed to users on the front end. Shadid stated that io.net’s system design allows for self-healing and continuous updates to each device to recover any incorrectly changed metadata. In response to this incident, io.net accelerated the deployment of user-level identity authentication integration with OKTA, which is expected to be completed within the next 6 hours. Additionally, io.net introduced Auth0 Tokens for user authentication to prevent unauthorized metadata changes. During the database recovery period, users will temporarily be unable to log in. However, all normal operating time records remain unaffected, and this incident will not affect computing rewards for suppliers.

4. Token Valuation

4.1 Token Model

The tokenomics of io.net involve an initial supply of 500 million IO tokens distributed across five categories: Seed Investors (12.5%), A-round Investors (10.2%), Core Contributors (11.3%), Research & Ecosystem (16%), and Community (50%). With the issuance of IO tokens to incentivize network growth and adoption, the total supply will increase to a fixed maximum of 800 million tokens over a period of 20 years.

The reward mechanism adopts a deflationary model, starting at 8% in the first year and decreasing by 1.02% monthly (approximately 12% annually) until reaching the 800 million IO token limit. As rewards are distributed, the shares of early supporters and core contributors will continue to decrease, with the community’s share growing to 50% once all reward allocations are completed. [4]

The utility of the IO token includes providing incentives for IO Workers, rewarding AI and ML deployment teams for continued network usage, balancing partial demand and supply, pricing IO Worker computing units, and facilitating community governance.

To mitigate payment issues arising from IO token price volatility, io.net has developed the stablecoin IOSD, pegged to the US dollar at a 1:1 ratio. 1 IOSD always equals 1 USD, and IOSD can only be obtained by destroying IO tokens. Additionally, io.net is considering implementing mechanisms to enhance network functionality. For example, IO Workers may be allowed to increase their chances of being leased by pledging native assets. In this scenario, the more assets they invest, the higher their probability of selection. Furthermore, AI engineers pledging native assets would have priority access to high-demand GPUs.

4.2 Token Mechanism

The IO token is primarily used for two major groups: demand-side and supply-side participants.

For demand-side participants, each computing job is priced in US dollars, and the network will hold the payment until the job is completed. Once node operators configure their reward shares in both US dollars and tokens, all US dollar amounts will be directly allocated to the node operators, while the portion allocated to tokens will be used to burn IO tokens. Then, during that period, all IO tokens minted as computing rewards will be distributed to users based on the dollar value of their coupon tokens (computing points).

For supply-side participants, rewards include availability rewards and computing rewards. Computing rewards are for jobs submitted to the network, where users can choose their preferred deployment time in hours and receive cost estimates from the io.net pricing oracle. Availability rewards involve the network randomly submitting small test jobs to evaluate which nodes run regularly and can effectively accept jobs from the demand side.

It’s worth mentioning that both supply-side and demand-side participants have a reputation system in place, accumulating scores based on computing performance and network participation to receive rewards or discounts.

In addition, io.net has ecosystem growth mechanisms, including staking, referral rewards, and network fees. IO token holders can choose to stake their tokens to node operators or users. Once staked, the stakers will receive 1-3% of all rewards earned by the participant. Users can also invite new network participants and share future partial income from these new participants. Network fees are set at 5%.

4.3 Valuation Analysis

Since we currently lack accurate revenue data from projects within the same sector, we cannot accurately estimate valuation. Therefore, our main point of reference will be a comparison with Render, a project similar to io.net in the AI+DePIN space, to provide some insights for consideration.

https://x.com/ionet/status/1777397552591294797

https://globalcoinresearch.com/2023/04/26/render-network-scaling-rendering-for-the-future/

As shown in the graphic, Render Network is currently the leading project in the AI+Web3 sector, focusing on decentralized GPU rendering solutions. It has a total of 11,946 GPU resources and a current market valuation of $3 billion (fully diluted valuation of $5 billion). On the other hand, io.net has a total of 461,772 GPU resources, which is 38 times more than Render. With both io.net and Render focusing on decentralized GPU computing as their core capability, it’s highly likely that io.net’s market valuation will surpass Render’s upon listing, or at least be comparable.

https://stats.renderfoundation.com/

Based on the provided data, Render Network had a Frames Rendered count of 9,420,335 and a GMV of $2,457,134 in 2022. Currently, Render Network’s Frames Rendered count has increased to 31,643,819, suggesting an approximate GMV of $8,253,751.

In comparison, io.net had a GMV of $400,000 in the first 4 months. Assuming io.net maintains this growth rate, the GMV for 12 months would reach $1,200,000. If io.net aims to achieve the current GMV of Render Network, it would need to grow approximately 6.8 times more.

Considering io.net’s potential, market competition, and the impact of a bullish market cycle, io.net has the potential to achieve a market valuation of over $5 billion during a bullish market cycle.

5. Summary

The emergence of io.net fills a gap in the decentralized computing field, providing users with a novel and promising computing approach. With the continuous development of fields such as artificial intelligence and machine learning, the demand for computing resources is constantly increasing, making io.net have high market potential and value.

On the other hand, although the market has given io.net a high valuation of $1 billion, its product has not yet been tested in the market, and there are uncertain risks in terms of technology. Additionally, whether it can effectively match its supply and demand relationship is also a key variable in determining whether its future market value can reach new highs. From the current situation, io.net’s platform has shown preliminary results on the supply side, but has not fully exerted its efforts on the demand side, leading to the overall GPU resources of the platform not being fully utilized. How to more effectively mobilize the demand for GPU resources is a challenge that the team must face.

If io.net can quickly integrate market demand and encounter no major risks or technical issues during operation, with its AI+DePIN’s tangible business attributes, its overall business will initiate a growth momentum and become one of the most prominent products in the Web3 field. This also means that io.net will be a very attractive investment target, so let’s continue to follow up, observe, and carefully verify.

Disclaimer：

1. This article is reprinted from [密客社区].转发原文标题《MIIX Capital: io.net项目研究报告》. All copyrights belong to the original author [密客社区]. 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.