A Research On IO.Net
Introduction
IO.Net stands at the forefront of decentralized physical infrastructure (DePIN) as a groundbreaking cloud computing platform. Its primary mission is to build the world’s largest AI computing DePIN, an enterprise-grade network that allows machine learning engineers to access distributed cloud clusters at a small fraction of the cost of comparable centralized services.
Since its inception in 2022, IO.Net has been making notable strides. A testament to its growing impact and adoption IO.Net boasts over 63K[1] GPU providers within its decentralized physical infrastructure network. The network’s capacity is impressive, featuring over 43K CPUs and 382K GPUs. This showcases the platform’s robust infrastructure and its potential to revolutionize the cloud computing landscape.
In terms of community management, IO.Net has attracted 475K followers on Twitter and 470K members on Discord, demonstrating strong support from the community. It’s evident that IO.Net is an exciting project at the intersection of decentralized computing and blockchain technology.
Leadership
At the helm of IO.Net is Ahmad Shadid[2], the founder and CEO.Shadid’s involvement in the tech industry has been extensive, with a focus on digital assets, API trading interfaces, and machine learning systems. During his time worked as a Quant Systems Engineer, he Invented DarkTick, a robust risk-management ML Powered engine that monitors and detects under-the-radar institutional money in-flows/out-flows by analyzing billions of trades tick-data” To Protect Investors capital from market crashes reflects his expertise in developing and managing complex systems.
Partnering with Ahmad Shadid in leadership is CMO Garrison Yang[3] who was previously Vice President of Growth and Strategy at Ava Labs Leading strategy and growth for Ava Labs’ consumer verticals (NFTs, Gaming, Enterprise, Institutions, DeFi) and product business units. And most of his experience focused is game realm.
COO at IO.Net is held by Tory Green[4]. Prior to joining IO.Net, he amassed almost a decade of experience at Tiller Partners, where he served as COO. During his tenure, Tory was instrumental in managing all investment functions, which included sourcing potential investments, conducting comprehensive market and company research.Tory also played a pivotal role in collaborating closely with founders of portfolio companies. He assisted in developing robust strategies for pricing, marketing, growth, recruiting, and operational enhancement. Additionally, Tory offered valuable advice and insights concerning exit strategies and opportunities for add-on acquisitions.With his extensive background in investment management, strategic planning, and operational expertise, Tory Green brings a wealth of knowledge and leadership to his role as COO at IO.Net.
Fundraising
On 5th,Mar,2024 IO.Net announced a $30million Series A round 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, as well as prominent industry leaders including Solana founder Anatoly Yakovenk, Aptos founders Mo Shaikh and Avery Ching, Yat Siu of Animoca Brands and Jin Kang of Perlone Capital[5], underscoring the investment community’s confidence in IO.Net’s vision.
Structure
IO Cloud
IO Cloud[6] is the place for costumers to deploy and manage on-demand decentralized GPU Clusters. It leverages a decentralized network of nodes (IO workers) to offer ML engineers and developers a seamless cloud-like experience. The platform integrates with IO-SDK, simplifying the scaling of AI and Python applications. IO Cloud’s core components are its clusters, which are self-healing and fully meshed GPU networks. This distributed system executes Python-based ML code efficiently and scales effortlessly to meet varying demands.
IO Worker
IO Worker[6] is designed to offer users a user-friendly webapp interface for effective management of GPU node operations. It includes features such as user account management, real-time monitoring of computing activities, temperature and power consumption tracking, installation assistance, wallet management, security measures, and profitability calculation.
IO Explore
IO Explorer [6] provides users with a comprehensive view of our GPU Cloud network, akin to a blockchain explorer like Solscan. Its purpose is to empower users to monitor and analyze every aspect of the IO.Net network, including statistics, data points, and Rewards transactions, with complete transparency and ease.
Backend
The backend of IO.Net encompasses the robust infrastructure and services that power its functionality. It includes server-side components such as databases, APIs, and computing resources that handle data processing, storage, and communication.
Data Performance
Worker Data
IO.Net’s recent achievements are truly impressive, with its decentralized physical infrastructure network now boasting over 63K GPU providers. This growth is complemented by a substantial capacity, including more than 42K CPUs and a staggering 370K GPUs.
The strategic partnership with Render and Fil last year bolstered these figures to 43K and 382K separately, a significant leap from its predecessor, Akash Network, which had only 281 GPUs and 19K CPUs. This exponential growth showcases IO.Net’s commitment to expanding its infrastructure and enhancing its computational capabilities.
Source: https://cloud.io.net/explorer/home
As of April 2, 2024
In summary, IO.Net’s remarkable progress in expanding its network, forging strategic alliances, and augmenting its computational resources positions it as a key player in decentralized infrastructure and AI-focused computing.
Clusters Data
We obtained cluster data from IO.Net and focused on the period from September 18, 2023, to April 2, 2024. It reveals that the number of clusters reached 4.2K, with a total payment of 487K USD and a net payment of 101K USD. This remarkable growth occurred in just six months, a stark comparison to Akash’s growth over two years. From the chart, we also acknowledge that the rate of cluster increase appears to be steady, with a slight acceleration in recent days. However, due to IO.Net still being in its test period for the mainnet, it faces a high refund rate.
Source: https://api.io.solutions/v1/io-explorer/clusters/{cluster_id}/details
As IO.Net progresses beyond its testing phase and stabilizes its mainnet operations, we may observe changes in refund rates and further optimizations in cluster growth and payments.
Idle Rate
IO.Net has achieved remarkable results in CPU and GPU supply, benefiting from its incentives for users who keep their devices online. However, the creation of only 21.5 clusters daily indicates that the project is still in its early stages and is experiencing a high idle rate. It’s important to note that this issue is not exclusive to IO.Net but is instead a broader challenge faced by the entire Depin community.
Lease Cost
We have chosen several centralized cloud computing platforms such as AWS, Azure, Google, Lambda Cloud, and two decentralized cloud computing platform, Akash Network, Vast AI for comparison.
Source: https://akash.network/gpus/
https://cloud.io.net/explorer/home
As of April 2, 2024
The table illustrates that IO.Net is less competitive regarding the H100, with users needing to pay 4 USD per hour, the highest rate among all platforms. However, IO.Net proves to be competitive with the A100, offering the lowest cost at 1.5 USD per hour. In the consumer-grade GPU segment, Akash demonstrates the highest overall price competitiveness, thanks to the introduction of their bidding mechanism “SuperCloud.” Following Akash, IO.Net ranks next, slightly ahead of Vast AI. Additionally, IO.Net provides better support for Apple chips.
Token Overview
IO is the native token of IO.Net. It serves as one of the few Coins out there with an actual use-case as a multi-layered governance coin. This token is built on the Solana blockchain, benefiting from its high throughput and low transaction costs. And we`ll see the final tokenomics until 15th,Apl,2024 after its release.
Token Allocation
Network Incentives: Reserved for rewarding GPU providers and network participants to encourage resource contribution.
Development Fund: Allocated for ongoing research, tech advancements, and network maintenance.
Team and Advisors: Set aside for the project’s team and advisors, often with a vesting schedule.
Ecosystem Growth: Dedicated to partnerships, marketing, and community-building initiatives.
Public Sale: Percentage available for sale to the public, promoting broader community involvement.
Token Utility
The utility of the IO token primarily revolves around governance, as officially confirmed by the team. It also plays a crucial role in securing the blockchain, incentivizing participants, and serving as a default mechanism for storing and exchanging value. Like some similar projects, IO token’s utility doesn’t extend to paying for GPU computing power due to its inherent volatility.
Conclusion
Advantages
Abundant Supply: Since its establishment in 2022, IO.Net has made significant progress, indicating its growing influence and widespread adoption. With over 63K GPU providers in its decentralized physical infrastructure network, IO.Net showcases a substantial presence in the cloud computing arena. The network’s capacity is equally impressive, boasting over 289K CPUs and 390K GPUs. These numbers highlight IO.Net’s robust infrastructure and its potential to transform the cloud computing landscape.
Competitive Pricing: Despite higher rates for certain resources like the H100, IO.Net maintains competitive pricing overall, especially with lower costs for other resources such as the A100. This flexibility allows users to choose cost-effective options based on their specific needs.
Specialized Services: IO.Net provides specialized services tailored to different hardware configurations, including support for Apple chips. This focus on diverse hardware compatibility ensures that users can leverage the platform’s capabilities regardless of their device preferences.
User-Friendly Interface: IO.Net offers a user-friendly interface that simplifies resource management, deployment, and monitoring. This intuitive interface contributes to a seamless user experience, making it easier for users to navigate and utilize the platform’s features effectively.
Challenges
Demand scarcity: The creation of only 21.5 clusters daily suggests that the project is still in its early stages and faces a high idle rate and poses a significant challenge for IO.Net. This challenge could impact the platform’s overall efficiency, resource utilization, and user experience. It’s crucial for IO.Net to address this issue effectively to ensure optimal performance and scalability in the long run.
Stability: The primary suppliers for IO.Net are individuals, unlike AWS, Azure, and Google Cloud, which have professional technicians to maintain their devices. This distinction can lead to perceived instability, potentially resulting in a less satisfactory experience for customers. We’ve seen this issue occur previously with Vast AI.
Competition: The competition in this field is highly intense, with giants like AWS, Azure, and Google Cloud dominating the centralized landscape. In the decentralized realm, there is competition from projects like Akash, Vast AI, and upcoming platforms such as Swan and AGI. Taking Akash for example, while IO.Net boasts a significantly higher supply than Akash, the overall pricing still seems somewhat lackluster, as revealed in the previous analysis and comparison. Therefore, IO.Net may need to make further adjustments to its pricing strategy to enhance its overall competitiveness and market appeal.
Finalthoughs
IO.Net signifies a notable advancement in the evolution of cloud computing, providing a decentralized, cost-effective, and technologically advanced alternative to conventional models. With strong leadership, robust community support, and ongoing technological innovations, IO.Net is well-positioned for substantial future growth.
Reference
[2] https://www.linkedin.com/in/grayspark/
[3] https://www.linkedin.com/in/garrisonyang/