Introduction

Spheron is a decentralized platform that taps into global computing resources, empowering users to flexibly and efficiently engage with advanced networks and deploy large-scale edge AI workloads. Through its Decentralized Compute Network (DCN), Spheron connects retail GPU providers and data centers to create a secure, seamless ecosystem that delivers the computational power users need.

Funding Background

Funding Details (source: rootdata）

Spheron Protocol has completed two rounds of funding. The seed round, held on April 19, 2021, raised $1.3 million, led by Ascensive Assets, with participation from notable investors such as SevenX Ventures, Master Ventures, and X21 Digital. A second funding round followed in August 2022, raising $7 million, with backing from Zee Prime Capital, Protocol Labs, ConsenSys Mesh, and Paradigm Shift Capital.

Team Members

Team Members（Source: rootdata）

The Spheron Network core team is led by Co-Founder and CEO Prashant Maurya and Co-Founder and CTO Mitrasish Mukherjee. Both have previously contributed to public blockchain projects such as LYNC and Bluzelle, although they do not have prior experience in the DePIN sector.

Core Features

Nodes

Spheron’s framework resembles DePIN networks like IO NET and Filecoin but introduces significant innovations. Resource providers can register as nodes in the Spheron network, offering GPU or CPU resources at customizable prices. The network supports two types of nodes that fuel the GPU leasing market: Master Nodes (Provider Nodes) and Light Nodes (Fizz Nodes).

Master Nodes (Provider Nodes)

At the heart of Spheron’s ecosystem are the provider nodes, which are essential for contributing CPU and GPU resources to the public resource pool. These nodes operate under a Kubernetes-based orchestration system and are incentivized through staking and rewards to contribute their computing power. This approach ensures the network remains robust and scalable.

Provider nodes form the backbone of the network, offering the CPU and GPU resources necessary to drive various applications. Leveraging Kubernetes, these nodes facilitate deployment orchestration, enhancing the network’s capability to scale based on different computational demands dynamically. This orchestration functionality simplifies deployment management and significantly improves the network’s overall efficiency and responsiveness.

Provider Registration and Staking (Source: docs.spheron.network）

Providers wishing to join the network must undergo a comprehensive verification process. This process ensures that only legitimate providers with sufficient computational resources and no malicious intent can join the network. The verification involves multiple steps, overseen by a Decentralized Autonomous Organization (DAO) or a governance body using a voting system.

1. Registration Proposal: Interested providers submit a proposal detailing their resources, including geographic location, computational capabilities, tier, and type of compute resources.
2. Governance Review: A governance body evaluates the proposal. Once approved, the provider’s information is registered in a provider registration smart contract. This process triggers the creation of a provider-specific smart contract and updates the proxy address in the registration contract.
3. Staking $SPON Tokens: Providers must stake $SPON tokens after registration. Staking activates the provider’s bidding engine, allowing them to participate in bidding for deployment tasks.

Light Nodes (Fizz Nodes)

Fizz Nodes are a lightweight version of provider nodes, specifically designed for operation on smaller devices such as personal computers or laptops. This innovation significantly boosts the network’s flexibility and accessibility, enabling community members to contribute their unused computing or GPU resources to the Spheron network. These nodes can collaborate to form subnets, consisting of a gateway node (or provider node) and several Fizz Nodes, allowing them to establish their own micro-economies and governance models.

Fizz Nodes rely on a Docker-based system, where each node operates in a dedicated Docker container based on user requests. Unlike resource-heavy solutions like Kubernetes, this approach minimizes computational demands on smaller devices. This architecture also addresses common challenges such as connectivity, isolation, security, and maintenance, ensuring the network remains decentralized and efficient.

Diagram of Fizz Node Architecture (Source: docs.spheron.network）

Key Components of the Fizz Node System:

1. Fizz Client: Operated by individuals on smaller devices, contributing underused computational resources to the network in exchange for financial rewards.
2. Gateway Service: Managed by providers or individual operators, this service facilitates interactions between users and Fizz Clients while handling most communication processes.

Deployment Lifecycle

The deployment lifecycle in a decentralized computing network is a streamlined process designed to allocate resources while maintaining security and transparency efficiently. The process starts with a user requesting server deployment and goes through several steps, including bidding, lease creation, activation, and potential updates or termination. Smart contracts power each stage to ensure fairness and reliability. Here’s how the lifecycle works:

• Starting and Finding Providers
  • The process begins when a user submits a server deployment request. This triggers a provider lock state, prompting all eligible providers in the network to compete by submitting bids to fulfill the request.
• Selecting Providers and Bidding
  • During the provider lock stage, providers that meet the deployment standards are invited to submit bids for the requested resources. The bidding period lasts 1-2 blocks, after which no further bids are accepted. The matching engine then processes the order, which selects the best provider based on the received bids. The results are forwarded to the order smart contract to initiate lease creation.
• Lease Creation and Provider Assignment
  • Once a provider is chosen, the smart contract updates the status to lease creation and prompts the provider to reserve the necessary resources. This step finalizes the agreement between the user and the provider via smart contract execution, ensuring resources are allocated per the request.
• Activating the Deployment
  • After confirming the reservation, the provider node updates the order contract to indicate readiness to receive the deployment manifest required for server setup.
  • The user then transfers the deployment manifest through an MTLS connection, enabling the provider to begin server deployment.
  • The provider configures and activates the server, ensuring it meets all specified requirements. Upon successful deployment, the provider updates the deployment status on-chain to deployment confirmation, including critical access details for the user, such as IP address, port, and connection URL.
• Termination and Shutdown
  • The user can terminate the deployment by issuing a lease closure event. Upon synchronization, the provider node will stop the server operation and mark the deployment as closed on the blockchain, ending the resource allocation.
• Updating Deployments
  • The lifecycle also accommodates modifications to existing deployments. Users can request updates to general configurations, specifications, or replicas through a lease update transaction. Once the provider synchronizes and implements the requested changes, the lease update event is recorded on-chain, reflecting the modifications and any related pricing adjustments.
  • At the core of the deployment process are smart contracts that automatically manage transactions, escrow funds, and rewards. These smart contracts ensure that the terms of deployment agreements are executed and that funds are transferred securely and efficiently based on resource usage. By leveraging blockchain technology, this system delivers transparency and establishes trust, which is crucial for maintaining the network’s integrity and earning user confidence.

Payment System

The Spheron Network has implemented an advanced payment system grounded in a token economy, ensuring transparent and fair rewards for GPU resource providers.

This payment system allows for using multiple tokens (including $SPON) as a medium of exchange, simplifying transactions and enabling seamless value transfers between users and providers. The system is designed to support a diverse ecosystem of contributors and allows organizations or participants to enhance the platform’s usability by enabling payments with various tokens. However, all tokens must undergo approval through a governance process before being accepted as payment. Built on blockchain technology, this payment system ensures that transactions are secure, auditable, and transparent, enhancing the network’s trust and reliability.

Below is a closer look at how the system works and the utility of the $SPON token:

User Payment Contributions

User Payments Workflow (Source: docs.spheron.network）

The workflow shown above illustrates that users must first deposit initial funds into Spheron’s escrow smart contract to initiate deployments within the protocol. This contract holds their balance (part of which is locked and part unlocked). Users can then create orders to begin new deployments via the order smart contract, which accesses the user’s unlocked balance to verify the minimum balance before starting the order-matching process.

Users must also specify the number of days/hours they wish to lock funds for the deployment order and the token they wish to use to compensate providers for their workload. Once the order matches a provider and the contract identifies the selected bid, the system calculates the initial funds to be locked for the order. After the lease is created, the funds are locked in escrow. Unless the user terminates the lease, they cannot access these funds. As the lease progresses, the locked amount is deducted from the user’s balance and allocated to the provider based on elapsed time. If the user closes the lease early, the user and provider’s balances are updated, and the final balances are distributed to their respective accounts, marking the lease as closed.

In summary, the system provides a transparent and flexible mechanism for user payments:

• Calculation of Escrow Wallet: Users set up a contract wallet on the blockchain as a computation escrow, where they can deposit the minimum amount. This escrow system allows users to deposit various tokens, including $SPON, enabling flexible fund management based on computational usage.
• Withdrawal and Fund Management: Users can withdraw any remaining funds from their escrow wallet, ensuring they pay only for the computational resources used. The escrow wallet also facilitates the transfer of instance ownership and funds between users, offering flexibility and control over resource allocation and usage.
• Maximum Resource Allocation Mechanism: After funds are deposited into the escrow wallet, a predefined bracketing system manages the allocation of computational resources. This system determines the amount of resources allocated based on the amount deposited, ensuring structured and fair distribution of computing power. For example, a $15 deposit entitles the user to specific computational resource allocations as follows:
  • For standard computational allocation: Access to 8 CPUs, 32 GB RAM, and 512 GB disk space.
  • For GPU-intensive tasks: Allocation includes the use of mid-range GPUs (e.g., A4000 or A6000 models), along with 32 CPUs, 128 GB RAM, and 1 TB of disk space.

Provider Compensation Mechanism

Order Matching for Providers (Source: docs.spheron.network）

When a user’s deployment order is successfully matched with a provider, a lease agreement is automatically created. This agreement locks a portion of the user’s funds, estimated to cover the deployment costs for the agreed duration (e.g., several days or months). Once the lease begins, these locked funds are gradually transferred to the provider as payment. While the lease is active, users cannot access the locked funds, but providers can withdraw the accumulated payments from their escrow accounts tied to the deployment.

Providers must also contribute 20% of their deployment earnings to the Spheron Foundation. If the lease is canceled by either the user or provider, both parties’ balances are updated, and the remaining funds are transferred to their respective accounts. Additionally, if the locked funds are insufficient to maintain an active lease, the lease smart contract will automatically terminate the agreement, and the system will trigger a shutdown of the server. To encourage the use of the native $SPON token and enhance economic efficiency within the network, Spheron employs a specific fee structure:

Usage Fees

• A 2% fee is applied to payments made with non-$SPON tokens.
• No fees are charged for payments made with $SPON tokens.

Provider Fees

• Active rewards do not incur any fees.
• For usage fees:
  • A 1% fee applies to payments made to providers, whether with $SPON or other tokens.
  • No fees are charged when providers are paid in $SPON tokens.

This fee structure incentivizes using $SPON tokens, reducing transaction costs for users and providers while supporting the network’s long-term sustainability.

Provider Compensation Model

Spheron’s provider compensation model is designed to ensure fairness and encourage participation:

• Usage Payments: Providers receive payments in the form of whatever token the user chooses to initiate the deployment. Providers can specify all tokens they wish to support and will be matched with orders paid in the tokens they support.
• Activity Rewards: If a provider’s node experiences low utilization during any Era period, the foundation will incentivize them to remain part of the network. This ensures that providers do not leave the network if they cannot generate sufficient returns from their hardware to cover operational costs. Rewards are issued in $SPON tokens and are distributed during specific periods known as Eras.

Development Prospects

The Expanding GPU Market（Source: blog.spheron.network）

The global GPU and data center market is experiencing rapid growth. According to Global Market Insights, the GPU market was valued at $52.1 billion in 2023, with a projected CAGR exceeding 27% from 2024 to 2032.

Projections indicate that GPUs will continue to grow in demand over the next few years due to the needs of artificial intelligence (AI) development, big data analytics, and cloud computing. Companies of all sizes rely on GPU-driven computation to handle complex tasks, train and operate machine learning models, and support the infrastructure powering modern applications. While the high-end market grabs most of the headlines, there is also a growing demand for low-end machines, particularly for testing, development, and non-production tasks.

Spheron caters to two primary audiences: AI developers who need flexible GPU access and Web3 developers seeking decentralized hosting and computation solutions. By meeting these diverse needs, Spheron positions itself as an essential tool for developers. It enables seamless scaling across different hardware layers, whether testing small AI models or launching advanced dApps. As the developer market grows, services that reduce complexity and lower costs will thrive. Spheron’s supercomputing model achieves both, enabling it to stand out in a crowded field.

Recent Developments

Spheron has launched a 5-phase plan to integrate 100,000 Fizzers into its network. These Fizzer nodes will form the backbone of Spheron, enabling AI at the edge and contributing to the network’s decision-making and operations. Each phase is named after celestial bodies in the solar system.

Recent Activities（Source: blog.spheron.network/join-the-fizz-phase-ii-powering-the-future-of-decentralized-ai)

• Phase 1 (5,000 users) - OG Fizzers
• Phase 2 (next 25,000 users) - Lunar Fizzer
• Phase 3 (next 50,000 users) - Stellar Fizzer
• Phase 4 (next 75,000 users) - Nebula Fizzer
• Phase 5 (final 100,000 users) - Galactic Fizzer

The project is currently in Phase 2. By joining the Fizz Node program, participants can contribute computing resources to the network while earning FN points as rewards.

Conclusion

AI computing and big data drive a growing need for scalable GPU solutions. The year 2024 has seen several DePIN projects, including io.net and Aethir, go live. The DePIN token sector has delivered higher returns than other segments, driven by the popularity of AI tokens. Across Web2 and Web3 ecosystems, AI remains a central focus. On-chain GPU leasing stands out for its flexibility, affordability, and robust security, creating a more efficient and user-friendly environment for providers and consumers of computing power.