Author: TechFlow
At a time when AI Agents are flooding the market and the code to wealth is becoming increasingly difficult to uncover, how can we more intelligently strike gold in the AI race?
From Solana to Base, from Pump.fun to Virtuals, the AI track is booming, with hundreds or even thousands of AI Agents being born every day. Its popularity is far beyond the reach of other tracks.
However, at present, most people's attention seems to be attracted by application-layer products that are more perceptible to the public, while the underlying infrastructure projects aimed at promoting higher-quality development of AI applications seem to be doing things quietly (wealth effect).
If you are already tired of the PvP battlefield of AI Meme, turning your attention from applications to the underlying layer may be a smart strategy for gold mining in the AI track. After all, the demand created by the prosperity on the application side will eventually be transmitted to the underlying layer, bringing immeasurable growth opportunities for AI infrastructure projects.
So, how will the journey to find Alpha from the bottom-level perspective of AI unfold?
Spheron breaks down the three key elements of AI development: data, models, and computing power to find market opportunities, aiming to integrate global computing resources to build a GPU trading market, and provide highly customized, low-cost computing services to computing power demanders including AI, DeFi, games and other applications, enabling the development of a decentralized economy and further entering the public eye.
What are the supply and demand challenges facing GPUs today? In a track like GPU that is strongly related to AI + DePIN, how does Spheron achieve peer-to-peer GPU supply and demand matching through the sophisticated design of Providers nodes and Fizz nodes and effective ecological incentives? As the AI boom continues and competition in the GPU track intensifies, what is Spheron's unique market position compared to other projects?
This article aims to take a closer look at Spheron and explore the huge potential of the decentralized computing power market and Spheron’s development opportunities by analyzing its track positioning and implementation logic.
Decentralization: The solution to GPU supply and demand imbalance
If you want to truly understand the subtlety of the "solution", the prerequisite is to be able to clarify the essence of the problem.
Why are GPUs important as a market that Spheron is targeting?
First of all, as an electronic circuit that can perform mathematical calculations at high speed, GPU has stronger parallel computing capabilities than CPU and can process large amounts of data streams simultaneously, making it an indispensable tool in the field of modern computing.
In fact, GPUs are already ubiquitous in our lives: whether it is graphics rendering and animation production for large-scale games, AI training and reasoning, or high-frequency trading and risk modeling in large-scale finance, GPUs provide a powerful driving force for building better products.
It can be said that as the wave of digitalization sweeps the world today, GPU has been regarded as an important cornerstone of the digital economy.
The importance of GPU is self-evident, so where does the problem in the GPU market arise?
Imbalance between supply and demand.
For GPU suppliers, on the one hand, the monopoly trend of GPU market supply is becoming increasingly obvious due to the high threshold for large-scale GPU production; on the other hand, the problem of GPU idleness is becoming increasingly prominent, and many GPUs in the market are not fully utilized, which not only causes waste of resources, but also affects the suppliers' profit returns.
For GPU demanders, the cost of obtaining GPUs is not only high, but also often cannot be flexibly adjusted according to their own needs. This undoubtedly compresses the space for most small and medium-sized enterprises to participate in market competition and further hinders the burst of innovation on the application side.
After clarifying the problem, what is the answer?
Over the past year, as AI, which has a strong demand for GPUs, has become popular in the decentralized Web3 market, market trends have given a choice.
An important advantage of the decentralized market is the low threshold for participation: anyone can contribute their idle computing power in a permissionless manner, or obtain computing power supply that better meets their actual needs. This openness allows more long-tail participants to enter the market, further expanding the scale of GPU market supply, while enhancing flexibility and promoting a more reasonable allocation of resources.
After establishing a connection between supply and demand, the greater advantage of the decentralized market is reflected in peer-to-peer transactions and efficient and convenient payment settlement: not only does it eliminate intermediaries and reduce transaction costs, but it also avoids the cumbersomeness and delay of traditional payment methods and realizes the efficient flow of bilateral market value.
Most importantly, well-designed token economics will provide a transparent profit distribution mechanism for the decentralized GPU market. The market can encourage wider market participation through token incentives, and each participant’s contribution will be honestly recorded and rewarded accordingly, bringing positive traction for the continued expansion of the ecosystem.
After exploring the background of the track, let's turn our attention to the specific logic of the solution:
Based on the power of decentralization, how does Spheron create a global GPU market where everyone can participate with low barriers and everyone can benefit fairly?
Global Supercomputing Network: How does Spheron achieve efficient matching of GPU supply and demand?
Simply put, Spheron's core architecture revolves around a matching engine, which aims to integrate global GPU resources and achieve efficient matching between the supply side and the demand side.
Computing power supply side:
Providers Node: Enterprise-level GPU provider, Spheron is the core pillar for building a global supercomputing network
Fizz Node: Integrate the long-idle home computing power, ordinary users can also contribute GPU
Computing power demand side:
Leading by various applications that require computing power
Provider nodes are closely related to the robustness and reliability of the network, so they have a high registration threshold and a strict review process:
Applicants first need to submit a registration proposal, which serves as a self-description of the GPU provider's own GPU capabilities, including specifications such as computing power, tier, and computing type;
Subsequently, the Spheron governance body will be responsible for verifying the proposal and deciding whether to pass it (in the initial operation phase of the network, all submitted proposals will pass after verification). In addition, Spheron has also developed multiple verification methods such as random verification and challenge mechanism to monitor the status of Providers nodes at all times and protect the network from fraudulent activities.
Introducing a staking mechanism for Providers nodes is an important measure for Spheron to further safeguard the rights and interests of the ecosystem: becoming a Providers node requires staking SPON tokens. If improper behavior occurs, the staked funds and income will be reduced or even confiscated, which aims to further constrain and incentivize GPU providers to maintain good behavioral standards.
Fizz nodes target ordinary users’ laptops and other hardware devices, and are committed to integrating the global long-tail computing power market with the lowest threshold, while further optimizing resource allocation, increasing broad participation and decentralization.
As an important part of the Spheron network, the Fizz node sets minimum CPU, RAM, storage capacity, and GPU support requirements, allowing ordinary laptops to easily contribute computing resources and gain additional benefits.
Participants first need to fill in relevant questions on the Spheron network to register a Fizz node. This process requires a small amount of ETH on the Spheron chain as Gas (currently available through the Spheron faucet or Arbitrum Sepolia in the testnet phase).
After completing the registration, participants can complete a simple setup by downloading the terminal and get up and running: Fizz client key components include pricing configuration, orchestrator, and service tunnel, which enable the underutilized computing resources of the device to be contributed to the network.
The gateway service is another core of the Fizz node, managed by the Providers node or personal gateway runner, and is designed to facilitate user interaction with the Fizz node client and handle most communications.
Based on the dual-node design of Providers and Fizz, the supercomputing network that Spheron is committed to building already has a basic framework. After aggregating resources, the bigger issue is how to use them efficiently. Spheron's powerful matching engine built on the AVS architecture plays an important role: the engine is built on Arbitrum and can quickly find the most matching GPU resources for both supply and demand parties based on region, hardware, and network speed. Users transmit deployment lists through MTLS connections to start workload deployment on designated providers.
First, Spheron adopts a grading mechanism for GPU suppliers, classifying them according to the capabilities and trustworthiness of different nodes, in order to achieve more accurate demand management and matching;
Spheron has a matching engine built on Arbitrum that can quickly find the most suitable GPU resources for both supply and demand parties based on region, hardware, and network speed. Users transmit deployment lists through MTLS connections to start workload deployment on designated providers.
When a GPU demander initiates a new deployment request, the supplier will monitor the demand and submit the bid to the smart contract;
After all bids are collected, the matching engine will select the solution that better meets the requirements of the demand side according to the predetermined parameters. The demand side will then verify the solution and send the deployment list to the matching supply side.
Finally, the GPU supplier consolidates the list and deploys the service, completing a more efficient, lower-cost and more adaptable demand matching.
Based on this, the framework for achieving efficient matching of GPU demand and supply has already come to light, and the effective ecological incentives brought by the ecological token SPON are the key to Spheron's stable and sustainable growth and continuous expansion of its ecology.
SPON Token: The finishing touch for Spheron to achieve a positive flywheel for the ecosystem
We all know that the economic model is one of the core designs of the blockchain network. It not only determines the function and use of tokens in the network, but also directly affects the long-term sustainability of the network, user incentive mechanism and the development direction of the ecosystem.
As the native token of Spheron, SPON has multiple utilities, aiming to promote the sustainable development of the ecosystem while providing diversified rights and benefits to token holders.
First, the Spheron network uses SPON tokens as Gas fees, which injects basic demand into the tokens.
Secondly, SPON tokens are the core value exchange medium in the Spheron payment system, providing users with efficient payment methods. Although the network supports other tokens as payment tools to enhance user flexibility and convenience, any non-SPON token payment method must be approved by the governance process to ensure the security and consistency of the network. Payment with SPON tokens is completely free, and other token payments require an additional 2% handling fee.
In addition, Spheron has introduced a staking mechanism for its tokens: on the one hand, Providers nodes need to stake SPON tokens to increase additional trust. When the nodes fail to fulfill their promises, their staked tokens will face the risk of being fined; on the other hand, token holders can also stake their SPON tokens to Providers nodes, thereby supporting the security and stable operation of the network and obtaining staking income.
One of the core functions of tokens is ecological incentives, and Spheron is no exception: whether it is a Providers node or a Fizz node, you can get SPON token rewards by deploying the node. In addition, in addition to earning income through staking, token holders can also get additional incentives by participating in ecological activities and supporting network development. This multi-level reward system ensures the enthusiasm of ecological participants.
In addition to being a payment tool and incentive medium, SPON tokens are also an important vehicle for ecosystem governance. Users holding SPON tokens can vote on major issues in the ecosystem, further ensuring ecological fairness and driving the network toward a direction that is more in line with the community's wishes.
Based on the SPON token incentive system, the valuable positive flywheel has been clearly seen in the Spheron ecosystem:
Relying on effective ecological incentives, more Providers nodes, Fizz nodes and demanders are attracted to join the ecosystem, which not only further improves the decentralization of the network, but also brings more active on-chain activities and revenue, making the network more capable of empowering its coin holders and bringing more substantial returns. This will also push the ecosystem into a spiral upward cycle, repeating itself and becoming stronger day by day.
In fact, in the more than one year since Spheron was born, the growing expansion of its ecosystem and the achievement of a number of breakthroughs have proven the rapid growth that this positive flywheel has brought to Spheron.
Mainnet is coming soon, tokens are about to be launched: Spheron has good future growth expectations
2024 is a very memorable year for Spheron. During this year, Spheron has achieved particularly remarkable ecological development.
Spheron's ecosystem has welcomed many important members in the past year: Spheron has established partnerships with Arbitrum, Caldera, Ora Protocol, Sentient, Witness, Genlayer, Heurist, DeNet, Akaave, Filecoin and many other top AI projects, becoming one of the important driving forces for the rapid development of Web3 AI.
Spheron also released a public beta network this year. According to the public beta data, the GPU utilization rate of the Spheron network is as high as 80%, which shows the efficient use of GPU resources under its powerful supply and demand matching engine.
In terms of nodes, after the successful launch of the Providers node, the community-focused Fizz node was also officially launched. More individuals can easily participate in computing power contributions in a low-threshold manner through devices such as laptops, and convert idle computing power into additional income.
In addition, Spheron's ecosystem includes the release of a console app that supports seamless GPU access, and the launch of a CLI and SDK to further lower the threshold for developers and help them easily integrate.
The achievement of a series of milestones has also brought about a rapid growth in the size of the Spheron community: According to the Spheron data dashboard, the total number of Spheron addresses has exceeded 36,000, and the number of connected nodes has exceeded 13,000, making it one of the largest DePIN computing networks currently.
By studying the Spheron project’s social media and published roadmap, more important project milestones are coming, injecting good expectations for continued growth in 2025.
In terms of ecological cooperation, Spheron recently announced its cooperation with Mira. The two will work together to develop trustless and scalable AI to further improve the accuracy, bias-free and credibility of AI output. Mira will also open 1,000 whitelist places specifically for the Spheron community. Users can apply using the code "SPHRN".
In addition, several major collaboration announcements including Witness, Genlayer, and Heuristic are in the works.
In terms of ecological functions, Spheron is about to launch the first service layer DePIN node Supernoderz, and will launch a model store to achieve seamless deployment of models.
More importantly, as the community voice continues to grow, the much-anticipated Spheron mainnet is about to arrive, and the token will be officially launched at that time.
As a key time node before the mainnet launch, users currently have multiple ways to efficiently participate in the Spheron ecosystem to obtain and accumulate future income chips. The most intuitive way is to deploy Fizz nodes. At the same time, the second phase of Lunar Fizzer activities has been launched, and 25,000 node operators will have the opportunity to obtain Lunar Fizzer status and exclusive rights. Specific rights include Lunar Fizzer status, airdrop priority, exclusive community activities, active rewards and rental rewards, etc., which have received widespread participation enthusiasm from the community.
The outstanding ecological performance also attracted the attention of industry organizations: In August 2022, Spheron revealed that it had received US$7 million in financing, and investors included Alphawave Ventures, NexusVP, Zee Prime Capital, Protocol Labs, ConsenSys Mesh, Paradigm Shift Capital, Matrix Partners India and Tykhe Ventures. Sandeep Nailwal, Aniket Jindal Julian Traversa, etc. According to community discussions, Spheron's new round of financing will be officially announced in the near future.
Conclusion
During the bonus period brought by the AI boom, the computing power market is undoubtedly one of the markets with the greatest potential and the most demand.
The GPU market that Spheron targets is hailed as the oil of the new era. Its sophisticated product design, effective ecological incentives, and a supply-demand matching mechanism that combines security, efficiency, cost, and flexibility will also enable various AI, gaming, and DeFi products that rely on GPUs to bring about a wave of decentralized innovation.
However, the AI craze has also brought about intensified competition in the GPU computing power track. Competitors include Aethir, Akash Network, Gensyn, Render Network, io.net, etc., and because the project logic is essentially to use tokens to incentivize computing power holders to participate in the network and provide computing power services, on the one hand, it improves the utilization rate of idle computing power, and on the other hand, it meets customers' computing power needs at a lower cost, achieving a win-win situation for both buyers and sellers. Therefore, the phenomenon of homogeneity and internal circulation in the track is becoming increasingly obvious.
In such a slightly crowded field, we can see that Spheron is emerging with its efficient and flexible features, thanks to its integration of the global computing market and a powerful supply and demand matching engine:
In terms of specific business, Render mainly serves 3D artists; io.net focuses more on AI and games; Aethir and Netmind focus on AI/ML engineers and enterprises;
In addition, since io.net's service model mainly focuses on leasing computing resources from nodes, although it has advantages in GPU resource leasing, it is slightly insufficient in terms of flexibility and adequacy in meeting customer needs. Render Network has the same problem. Although it improves rendering efficiency through its off-chain rendering function, it still lacks flexibility and adaptability to different customer needs due to its reliance on off-chain rendering.
Spheron aims to achieve computing power matching that is more in line with actual needs through Providers nodes, Fizz nodes, grading mechanisms, and a powerful matching engine, bringing more flexible GPU services that can better meet the needs of diverse users. It will also enable various AI, games, and DeFi products that rely on GPUs to bring a wave of decentralized innovation.
With the launch of Fizz nodes and the arrival of the mainnet and tokens, for investors who are concerned about the GPU computing power market, perhaps personally participating in the node deployment and experiencing the product details can give a clearer judgment than simple text.