Author: MIIX Capital Source: Medium

1. Project situation

1.1 Business summary

IO.NET is a decentralized GPU network, which aims to provide calculations for ML (machine learning).By assembling more than 1 million GPUs from independent data centers, cryptocurrency miners, and Filecoin or Render to obtain computing power.

Its goal is to combine 1 million GPUs into DEPIN (decentralized physical infrastructure network) to create an enterprise -level and decentralized distributed computing network.), Provide artificial intelligence engineers with lower price, easier gains, and more flexible adaptation of online computing resource services.

For users, it is equivalent to a bazaar that goes to the world’s idle GPU resources, allowing artificial intelligence engineers or teams to customize and purchase the GPU computing service required for purchasing according to their needs.

1.2 Team background

>

Ahmad Shadid is the founder and CEO, and was previously an engineer of Whalestrader quantitative system.

Garrison Yang is the chief strategic officer and chief marketing officer, and was previously the vice president of AVA LABS growth and strategic.

Tory Green is the chief operating officer and was previously the chief operating officer of Hum Capital, the Fox Mobile Group enterprise development and strategic director.

Angela Yi is the vice president of business development and graduated from Harvard University in the United States, responsible for planning and implementing key strategies such as sales, partnership and supplier management.

When Ahmad Shadid built the GPU computing network for the machine learning quantitative trading company Dark Tick in 2020, because the transaction strategy is close to high -frequency trading, a large number of computing power is required. The high GPU service cost of cloud service manufacturers has become their problem.

The huge demand for computing power and the high costs facing them prompted them to decide to do decentralized distributed computing resources, and then got attention in Austin Solana Hacker House.Therefore, IO.NET belongs to the team from the pain point faced by itself, proposing solutions and conducting business implementation and expansion.

1.3 Product/Technology

>

Problems faced by market users:

It is limited availability. Using cloud services such as AWS, GCP, or Azure usually takes a few weeks, and the GPU model popular on the market is usually unavailable.

There are very few choices, such as the GPU hardware, location, safety level, delay and other users have almost no choice.

High cost: It is very expensive to get high -quality GPUs, and it cost hundreds of thousands of dollars per month for training and reasoning.

Solution:

GPUs that are not fully utilized (such as independent data centers, encryption miners, and encrypted projects such as Filecoin, Render), integrate these resources into DEPIN, so that engineers can obtain a lot of computing power in the system.It allows the ML team to cross -distribute GPU networks to build reasoning and model service work processes, and use distributed computing libraries to arrange and batch training operations in order to use data and model parallelism on many distributed devices.

In addition, IO.NET uses a distributed computing library with advanced superal adjustment to check the best results, optimize scheduling, and simply specify the search mode.It also uses an open source reinforcement learning library that supports production -level and highly distributed RL (reinforcement learning) workloads and simple APIs.

Product composition:

IO Cloud, the purpose is to deploy and manage the decentralized GPU cluster as needed, seamlessly integrated with IO-SDK, and provides a comprehensive solution to expand artificial intelligence and Python applications.It can provide unlimited computing power, while simplifying the deployment and management of GPU/CPU resources.

IO Worker provides users with a comprehensive and user -friendly interface, and efficiently manage their GPU node operations through intuitive network applications.The scope of this product includes functions related to user account management, computing activities monitoring, real -time data display, temperature and power consumption tracking, installation assistance, wallet management, security measures, and profitability.

IO Explorer mainly provides users with comprehensive statistical data and visual visualized diagrams in all aspects of GPU cloud, allowing users to easily monitor, analyze and understand the complex details of the IO.NET network, and provide online activities, important statistics, data points and reward transactionsComprehensive visibility.

Features:

Decentralization computing network: IO.NET uses decentralized computing mode to distribute computing resources around the world, thereby improving the efficiency and stability of computing.

Low -cost visit: Compared with traditional centralized services, the IO.NET Cloud provides lower access costs, enabling more machine learning engineers and researchers to get computing resources.

Distributed gathering group: The platform provides a distributed gathering cluster. Users can choose appropriate computing resources according to their needs and allocate tasks to different nodes for processing.

Support machine learning tasks: IO.NET Cloud focuses on providing computing resources for machine learning engineers, so that they can make tasks such as model training and data processing easier.

1.4 Development Roadmap

>

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

According to the information published by IO.NET White Paper, the route map of the project product is: January-April 2024, V1.0 was fully released, dedicated to the decentralization of the IO.NET ecosystem, socopy.

1.5 financing information

>

According to public news information, on March 5, 2024, IO.NET announced the completion of $ 30 million in Series A financing, Hack VC led, Multicoin Capital, 6th Man Ventures, M13, Delphi Digital, Solana Labs, Aptos Labs, ForeSightVentures, Longhash, Sevenx, Arkstream, Animoca Brands, Continue Capital, MH Ventures, Sandbox Games and other participation.[1] It is worth noting that after this round of financing, the overall valuation of IO.NET is $ 1 billion.

2. Market data

2.1 Official website

>

>

From the official website data from January 2024 to March 2024, the total number of visits is 5.212m, the average monthly access is 1.737m, and the output rate is 18.61%(low).Search access accounts for more than 80%, which may indicate that the proportion of dirty data in the user data is not high. They have a basic understanding of IO.NET and are willing to learn more about interaction with the website.

2.2 Social media community

>

3. Competitive analysis

3.1 Competitive pattern

The core business of IO.NET is related to the central AI computing power. Its biggest competitors are traditional cloud service vendors represented by AWS, Google Cloud, and Microsoft Smart Cloud Business (Azure).According to the “2022 -2023 Global Computing Power Index Evaluation Report” by the International Data Company (IDC), Inspur Information and Tsinghua University Global Industrial Research Institute, the global artificial intelligence computing market is expected$ 34.66 billion in 2026.【2】

Compared with the sales revenue of the world’s mainstream cloud computing manufacturers: In 2023, AWS cloud service sales revenue was US $ 9.08 billion, Google Cloud’s sales revenue was 3.37 billion US dollars, and Microsoft Smart Cloud business sales revenue was US $ 9.68 billion.[3] The market share of the three accounts for about 66%of the world’s market share. At the same time, the market value of the three giant companies is more than trillion US dollars.

>

https://www.alluxio.io/blog/maximize-gpu-Utilization-For-model-traw

In stark contrast to the high income of cloud service manufacturers, how to increase the utilization of GPUs has become a focus problem.According to a survey of AI Infrastrus, most GPU resources are underestimated and utilized -53%of people think 51 ~ 70%of GPU resources are undervalued and utilized. 25%believe that the utilization rate reaches 85%, only 7%of 7%People believe that the utilization rate exceeds 85%.For IO.NET, the huge demand for cloud computing and the lack of effective use of GPU resources are the market opportunities.

3.2 advantage analysis

>

◢ @ionet_official Bull Stats

→ Accrued > $400k in the last four months

→ Clustered +40k GPUs and +5600 CPUs (including @rendernetwork and @Filecoin )

→ +69K worker nodes and continuously increasing

→ Deploy 10.000 GPUs without geographical limitation (+70 countries)… pic.twitter.com/KcotyyAumK

— Eli5DeFi (@eli5_defi) March 14, 2024

IO.NET’s biggest competitive advantage is reflected in the advantages of ecological position or first -mover advantage.According to the official data: At present, the total amount of GPU clusters owned by IO.NET is greater than 40K, the total amount of CPU is greater than 5600, Woker Nodes is greater than 69K, the time of deployment of 10,000GPU is less than 90s, the price is 90%cheaper than competitors, and the valuation is 1 billion.Dollar.IO.NET not only provides customers with a low price of a centralized cloud service provider at a low price and instant online service without permission. It also provides additional startups through the upcoming IO tokens.Inspiration, jointly help to achieve the goal of connecting 1 million GPUs.

In addition, compared with other DEPIN computing items, IO.NET focuses on the GPU computing power. The size of its GPU network has led over 100 times ahead of similar projects.IO.NET is the first of the most advanced ML technology stack (such as the RAY cluster, Kubernetes cluster and giant cluster) of the most advanced ML technology stack into the GPU DEPIN project and put it into large -scale practice. This makes it not only in the number of GPUs, but alsoThe ability of technology applications and model training is a leading position.

With the continuous development of IO.NET, if the GPU capacity can be increased to 500,000 full networks competing with centralized cloud service providers, the GPU will be available, which will provide similar services to Web 2 with lower costs and have the opportunity to pass the opportunity to passThe close cooperative relationship established by the main DEPIN and AI players (including Render Network, Filecoin, Solana, Ritual, etc.) gradually established its core position in this field to become the leader and settlement layer of the decentralized GPU network.Bring vitality.

3.3 Risk and question

IO.NET is an emerging computing resource integration and distribution platform that combines the in -depth combination of Web3, and the business involved with the traditional cloud service manufacturer has a high degree.Essence

As a emerging platform, IO.NET has not experienced large -scale application tests, nor does it reflect the ability to prevent and deal with malicious attacks.In the face of huge amounts of computing power resource access, distribution, and management, there is no corresponding experience or practical verification, and it is easy to occur with problems such as compatibility, robustness, and safety of technical products.And once problems occur, it is likely to be fatal to IO.NET, because customers are more concerned about their safety and stability, and they are unwilling to pay for these.

The market expansion is slow. IO.NET is highly reorganized with the traditional cloud service manufacturers, which allows it to compete directly with the traditional AWS, Google Cloud, Alicloud and other direct competition, and even compete directly with the second or third -line service providers. Although IO.NET has moreThe cost of discounts, but it has just begun to the service system and market system for Class B customers, which is very different from the market operations of the existing Web3 industry. Therefore, it is not currently the progress of market expansion in terms of market expansion.Ideal, this is likely to directly affect its project valuation and the market value of tokens.

Latest security incident

On April 25th, IO.NET founder and CEO AHMAD Shadid was pushed that IO.NET metadata API encountered security incidents. The attacker used the accessable mapping of the user ID to the device IDThis vulnerability does not affect the GPU access, but it does affect the metadata displayed by the front end to the user.IO.NET does not collect any PII, nor will it leak sensitive user or device data.

Shadid said that the design of the IO.NET system allows self -repair, constantly updates each device to help restore any error -changing metadata.In view of this incident, IO.NET has accelerated OKTA’s user -level authentication and integration deployment, which will be completed within 6 hours.In addition, IO.NET also launched Auth0 Token for user verification to prevent unauthorized metadata change.During the database recovery, users will be unable to log in for the time being.所有正常运行时间记录均不受影响，并且这不会影响供应商的计算奖励。

4. To token valuation

4.1 token model

>

IO.NET token economic model will have a initial supply of 500 million IO when the creation is created, divided into five categories: seed investors (12.5%), Series A investor (10.2%), and core contributors (11.3%), R & D and ecosystems (16%) and communities (50%).With the issuance of IO to motivate network growth and adoption, it will increase to a fixed maximum supply of 800 million pieces in 20 years.

The reward adopts a shrinking model, starting from 8%in the first year, a monthly decrease of 1.02%(about 12%per year) until the upper limit of 800 million IO.With the distribution, the share of early supporters and core contributors will continue to decrease. After all the reward distribution is completed, the community share will increase to 50%.【4】

Its token function includes IO Worker allocation incentives, rewarding AI and ML deployment teams to continuously use networks, balanced parts and supply, pricing for IO worker computing unit, and community governance.

In order to avoid the problem of payment caused by the fluctuations of the IO currency price, IO.NET is specially linked to the development of stablecoin iOSD.1iosd is always equal to $ 1.IOSD can only be obtained by destroying IO.In addition, IO.NET is considering some mechanisms to improve network functions.For example, IO workers may be allowed to increase the probability of being rented by mortgage native assets.In this case, the more assets they invest, the greater the probability they choose.In addition, artificial intelligence engineers with pledged assets can give priority to high -demand GPUs.

4.2 tokens mechanism

The IO tokens are mainly used for the two major groups of demand parties and suppliers. For demand parties, each computing operation is priced in US dollars, and the network will be retained until the operation is completed.Once node operators allocate their reward shares with US dollars and tokens, all US dollars will be directly assigned to node operators, and the share to tokens will be used to burn IO coins.然后，在该期间作为计算奖励铸造的所有 IO 币都会根据其优惠券代币（计算积分）的美元价值分配给用户。

For suppliers, including availability rewards and computing rewards.Among them, the calculation reward is for operations submitted to the network. Users can choose time preferences “the duration of the cluster in the hour” and the cost of receiving the cost of receiving the cost of the pricing prophet from the IO.NET pricing machine.In terms of availability rewards, the Internet will randomly submit small test operations to evaluate which nodes run regularly and can accept homework from the demand side.

It is worth mentioning that, both the supplier or the demander, there is a set of reputation systems, which accumulate scores based on computing performance and network participation to get rewards or discounts.

In addition, IO.NET also set up an ecological growth mechanism, including pledge, invitation rewards and network costs.IO currency holders can choose to pledge its token IO to node operators or users.Once pledged, the pledker will get 1–3%of all the rewards obtained by the participants.Users can also invite new network participants and share part of the future income of new participants.The network cost is set up 5%.

4.3 Valuation Analysis

At present, we cannot get accurate income data from the projects in the track, so we cannot accurately valuation. Here we mainly compare the project Render with the same AI+DEPIN project with IO.NET for your reference.

>

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

>

https://globalCoinresearch.com/2023/04/26/Render-sCaling-rendering-for-future/

As shown in the figure, Render Network is currently the leading project of decentralized GPU rendering solution in the AI+Web3 track. The total amount of GPU resources is 11946, and the current market value is US $ 3 billion (FDV5 billion US dollars); and the total amount of IO.NET GPU resources461772, 38 times the render, currently valuation of 1 billion.For the IO.NET and RENDER projects, the core key capabilities of the two are decentralized GPU computing power. Therefore, from the perspective of the GPU supply as the core comparison dimension, the market value of IO.NET’s market value will exceed the render, at least at leastNot comparable.

>

https://stats.renderfoundation.com/

Render Network’s Frames Rendered in 2022 is 9,420,335, and GMV is $ 2,457,134. At present, the Frames rendered of Render Network is 31,643,819, thereby calculating the entire GMV of about $ 8,253,751.

In contrast IO.NET 4 months GMV is 400,000, assume that IO.NET increased at the average speed of 4 months at GMV400,000, and 12 -month GMV is 1200,000.6.8 times the growth space, now IO.NET is valued at US $ 1 billion. Based on the above analysis, the market value of IO.NET in the bull market cycle is expected to reach more than $ 5 billion.

5. Summary

The emergence of IO.NET fills the gap in the field of decentralization and calculation, providing users with a novel and potential calculation method.With the continuous development of artificial intelligence and machine learning, the demand for computing resources is also increasing, so IO.NET has high market potential and value.

On the other hand, although the market has given IO.NET’s high valuation pricing of $ 1 billion, its products have not been tested by the market, and there are uncertain risks in terms of technology, and whether it can effectively match its supply and demand relationship is also determined that it is also determined that it is also determined that it is also determined that it is also determined that it is also determined that it is also determined that it is also determined that it is also determined that it is also determined that it is also determined that it is also determined that it is also determined that it is also determined that it is determined that it is also determined that it is also determined that it is also determined thatWhether its follow -up market value can reach a new high variable.Judging from the existing situation, the results of the IO.NET platform on the supply side have been initially emerged, but the demand side has not fully exerted its strength, resulting in the overall GPU resources of the current platform.The demand for resources is a challenge to the team.

If IO.NET can complete the fast access of market demand and does not encounter or major risk and technical problems during the operation, with its physical business attributes of AI+DEPINThe most eye -catching project products in the Web3 field also means that IO.NET will be a high -quality investment target of sub -factories, let us continue to follow up and observe and carefully verify.