← Back to Homepage

UNBY Cloud Computing Power Digital Currency Whitepaper

Version: V1.0

UNBY is a decentralized cloud computing network built for the global compute economy, dedicated to converging, scheduling, and efficiently flowing heterogeneous computing resources distributed worldwide — making computing power a universally accessible infrastructure for the digital economy. This whitepaper systematically presents UNBY's vision and mission, technical architecture, ecosystem, use cases, governance mechanism, and development roadmap, providing a comprehensive project framework for investors, the technical community, and industry partners.

Table of Contents

  • Abstract
  • Vision & Mission
  • Industry Background & Market Opportunities
  • Cloud Computing Power Concept
  • System Architecture
  • Ecosystem
  • Use Cases
  • Governance Mechanism
  • Development Roadmap
  • Team & Advisors

1. Abstract

Computing power is becoming the core productive force of the digital economy. As demand for compute grows exponentially across AI model training, scientific computing, 3D rendering, and edge computing, the global supply-demand imbalance is becoming increasingly acute: on one hand, massive compute demand cannot be met in time; on the other, vast amounts of distributed computing resources remain idle due to the lack of efficient scheduling mechanisms.

UNBY proposes a cloud computing power framework that, by building a decentralized distributed computing network, integrates data centers, GPU clusters, edge devices, and even personal idle compute worldwide into a unified computing resource pool, enabling on-demand scheduling, precise metering, and efficient flow of computing power.

The core philosophy of UNBY is "making compute flow like water" — anyone who owns computing power can become a provider, anyone who needs computing power can easily access it, and the supply and demand of compute are freely matched in an open, transparent, and efficient market.

This whitepaper is organized around the following core themes:

  • Industry Insights: Analyzing the explosive growth trend of global compute demand and current market pain points
  • Conceptual Framework: Systematically presenting the theoretical foundations of cloud computing power and compute assetization
  • System Architecture: Detailing how the four-layer technical architecture supports efficient scheduling and circulation of compute
  • Ecosystem: Depicting the multi-party value network and ecosystem flywheel effect
  • Use Cases: Deep analysis of solutions and value creation across four core application domains
  • Governance & Roadmap: Articulating community governance mechanisms and a phased development blueprint

UNBY believes that the democratization and assetization of computing power will reshape the resource allocation paradigm of the digital economy, unleash tremendous social productivity, and transform computing resources from the exclusive facilities of a few giants into shared digital infrastructure for all of humanity.

2. Vision & Mission

2.1 Vision: Making Compute Flow Like Water

In the industrial age, electricity was the core energy driving society; in the digital age, computing power has become the core productive force driving intelligent society. Yet today's computing is far from achieving the universal flow that electricity and water have achieved — computing resources are highly concentrated in the hands of a few large cloud providers, with high access barriers, low scheduling efficiency, and insufficient resource utilization.

UNBY's vision is to build an open, inclusive, and efficient global computing network, making computing power a freely flowing digital infrastructure. In this network:

  • Accessing compute is as convenient as turning on a faucet
  • Supplying compute is as flexible as connecting to a power grid
  • Flowing compute is as efficient as resource circulation
  • Metering compute is as precise and transparent as reading a utility meter

We believe that when computing power truly flows freely, it will unleash social productivity far beyond what we can imagine today, enabling every innovator — regardless of location or scale — to equally access computing resources and participate in building the digital economy.

2.2 Mission: Building the Trust Infrastructure for the Compute Economy

UNBY's mission is to build the trust infrastructure connecting compute supply and demand, using technology to eliminate information asymmetry, reduce transaction friction, and establish a fair, transparent, and efficient market order for computing.

Specifically, UNBY is committed to:

  • Aggregating globally distributed compute: Bringing data centers, GPU clusters, edge devices, and personal idle compute into a unified resource pool
  • Enabling intelligent scheduling and matching: Precisely matching compute supply and demand based on multi-dimensional metrics
  • Establishing a trusted metering system: Measuring compute contributions and consumption with standardized metrics, ensuring fairness and transparency
  • Advancing compute assetization: Transforming compute from a consumed resource into a measurable, tradable digital asset

2.3 Three Pillars

UNBY's vision and mission rest on three core pillars:

Pillar One: Open & Inclusive

  • Breaking the computing monopoly, enabling small and medium institutions and individual developers to access high-quality compute at reasonable cost
  • Lowering the barriers to compute access — any qualified compute provider can participate in the network
  • Extending computing resources to underdeveloped regions, narrowing the digital divide

Pillar Two: Efficient Flow

  • Building an intelligent scheduling engine to achieve dynamic optimal allocation of cross-regional, cross-architecture computing resources
  • Establishing a standardized compute measurement system, making different types of computing comparable and tradable
  • Shortening the compute supply-demand matching chain, reducing intermediary overhead

Pillar Three: Sustainable Development

  • Prioritizing the integration of renewable-energy-powered computing resources, reducing carbon footprint
  • Reducing resource waste from redundant construction by improving compute utilization
  • Advancing green computing technology innovation, building an environmentally friendly compute ecosystem

These three pillars mutually reinforce and work in concert, collectively forming the foundation of the UNBY computing network. Open and inclusive is the value orientation, efficient flow is the technical guarantee, and sustainable development is the long-term commitment. UNBY will use these as its core, continuously driving the healthy evolution of the global compute economy ecosystem.

3. Industry Background & Market Opportunities

3.1 Explosive Growth in Global Compute Demand

The world is in a historic period of explosive compute demand. Multiple technology waves are converging, collectively driving unprecedented demand for computing resources.

The global cloud computing market continues to grow rapidly, with an expected compound annual growth rate exceeding 20% over the next 5 years.
Global data center compute demand is growing at over 30% annually.

The AI wave is currently the most powerful engine for compute demand growth. Large language models (LLMs), multimodal models, recommendation systems, and other AI applications require exponentially increasing compute for training and inference:

AI large model training compute demand is growing exponentially, doubling every 3-4 months.

This means the rate of compute demand growth has far exceeded the pace of hardware performance improvement described by Moore's Law, and the supply-demand gap continues to widen.

Scientific computing also contributes substantial compute demand. Genome sequencing, drug discovery, climate modeling, astrophysics, and other frontier research rely heavily on high-performance computing resources.

3D rendering and creative industries are seeing continuously rising demand for rendering compute as film VFX, game development, and architectural visualization undergo digital upgrades.

Edge computing, as an important complement to cloud computing, is rapidly being deployed in smart city, autonomous driving, and industrial IoT scenarios:

The edge computing market is expanding rapidly, expected to reach hundreds of billions of dollars by 2027.

3.2 National Strategic Initiatives

Computing power has become an important indicator of national competitiveness. China's "East Data West Computing" project is a major strategic deployment of the national computing network, building a unified national computing network to guide computing demand from the east to renewable-energy-rich western regions, achieving cross-regional optimal allocation of computing resources. This strategy aligns closely with UNBY's vision of efficient compute flow and green sustainable development.

Internationally, ITU-T (International Telecommunication Union Telecommunication Standardization Sector) has published computing network-related standard frameworks, and ISO/IEC has established series standards in cloud computing and distributed computing. This indicates that computing networks and computing infrastructure have become a global consensus, providing policy and standards-level support for the realization of UNBY's vision.

3.3 Current Compute Market Pain Points

Despite strong compute demand, the current compute market has numerous structural pain points that severely constrain the effective utilization of computing resources:

Pain Point One: Severely Unequal Resource Distribution

Computing resources are highly concentrated at both the institutional and geographic levels. Large tech companies and hyperscale data centers control the vast majority of high-end computing, while many small and medium institutions, research teams, and individual developers face difficulties in accessing compute. Geographically, computing resources are concentrated in a few developed regions, with severely insufficient supply in underdeveloped areas.

Pain Point Two: Low Resource Utilization

Due to the lack of effective global scheduling mechanisms, a significant proportion of deployed computing resources are idle or under low load. Industry estimates suggest that a considerable portion of server compute is underutilized during off-peak hours, and the idle rate of personal devices and small computing facilities is even higher. If these "dormant" computing resources could be effectively integrated, they would form a substantial incremental supply.

Pain Point Three: High Access Costs

Traditional cloud computing services adopt a centralized supply model with many intermediaries and opaque pricing mechanisms, resulting in persistently high costs for end users to access compute. For small and medium innovation teams, high computing costs have become a significant bottleneck constraining their technical exploration and product iteration.

Pain Point Four: Severe Carbon Emission Pressure

Data center energy consumption accounts for approximately 1-2% of global electricity usage.

As compute demand continues to rise, data center energy consumption and carbon emissions are becoming increasingly prominent. Traditional centralized computing infrastructure relies heavily on fossil fuels, creating tension with global carbon neutrality goals. PUE (Power Usage Effectiveness) has become a key metric for measuring data center energy efficiency, with the industry universally pursuing lower PUE values.

3.4 Market Opportunities

The above pain points also harbor tremendous market opportunities:

  • Idle compute activation: If the vast amount of globally idle computing resources can be effectively integrated, it would form a substantial incremental supply with enormous market potential
  • Decentralized supply: The distributed computing network model can significantly reduce intermediary costs, providing users with more cost-effective compute access
  • Green compute premium: Renewable-energy-powered computing resources will gain market favor, with green computing becoming a differentiated competitive advantage
  • Compute assetization: The transformation of compute from a consumable to a tradable asset will spawn entirely new business models and economic ecosystems
  • Inclusive compute services: There is a significant market gap for inclusive computing services targeting small and medium institutions and developers

UNBY was born precisely at this historic opportunity, using decentralized concepts and technical architecture to directly address the structural pain points of the compute market and open a new paradigm for the compute economy.

4. Cloud Computing Power Concept

4.1 Definition and Connotation of Computing Power

Computing Power, broadly defined, refers to the capability of computing devices or systems to perform information processing and computational tasks. In the digital economy era, computing power has become a core productive factor alongside electricity and thermal energy, serving as the foundational support for all digital applications including AI, big data, and IoT.

Computing power is not a single-dimensional concept. Based on the characteristics of computational tasks, it can be categorized into several types:

  • General computing: General-purpose computing capability represented by CPUs, suitable for logic control and serial computing tasks
  • Intelligent computing: Parallel computing capability represented by GPUs and AI accelerators, suitable for deep learning training and inference
  • Edge computing: Lightweight computing capability deployed at the network edge, suitable for real-time processing in low-latency scenarios
  • Heterogeneous computing: Comprehensive computing capability with multiple computing units (CPU, GPU, AI accelerators) working in coordination

Measuring computing power also requires a multi-dimensional metric system:

Compute metrics include FLOPS (floating-point operations per second), throughput, latency, and other multi-dimensional indicators.

A single metric cannot comprehensively reflect the true capability of computing. UNBY adopts multi-dimensional metrics to comprehensively evaluate the quality and suitability of computing resources.

4.2 The Concept of Cloud Computing Power

Cloud computing power is the deepening and extension of the cloud computing concept in the computing domain. Traditional cloud computing provides a "rental" model for computing services, where users rent fixed-specification computing instances from cloud providers. Cloud computing power further abstracts compute from a "service" to a measurable, schedulable, tradable "resource element," emphasizing the networked aggregation and global optimization of computing.

Core characteristics of cloud computing power include:

  • Resource pooling: Aggregating geographically dispersed, architecturally heterogeneous computing resources into a unified logical resource pool
  • On-demand scheduling: Dynamically allocating computing resources based on task demand, achieving real-time matching of supply and demand
  • Elastic scaling: Computing supply automatically expanding or contracting with demand changes, avoiding resource waste
  • Decentralization: A distributed network architecture that does not rely on a single centralized platform, with multi-party participation
The Computing Power Network concept has been proposed by institutions such as China's CAICT, emphasizing computing as infrastructure.

The vision of the computing power network is to elevate computing to a national and societal level infrastructure — just as the power grid is to electricity and the communication network is to information — the computing power network will become the new infrastructure of the digital economy era.

4.3 The Necessity and Significance of Compute Assetization

Compute assetization is the process of transforming computing power from a purely consumed resource into a measurable, tradable digital asset. This transformation carries profound necessity and significance:

Necessity

  • Solving the measurement challenge: As an intangible resource, computing power has long lacked a unified, widely accepted measurement standard. Assetization requires establishing a standardized measurement system, making compute contributions quantifiable and comparable
  • Incentivizing supply willingness: When compute contributions can be precisely measured and rewarded with corresponding value, more compute owners will be incentivized to participate in the network, releasing idle compute
  • Promoting circulation efficiency: Assetized compute has divisible and transferable characteristics, making compute supply-demand matching more flexible and efficient

Significance

  • Reshaping resource allocation: Compute assetization transforms computing resources from the closed facilities of a few giants into circulating elements in an open market, driving resources to flow to the most efficient users
  • Spawning new economic ecosystems: As a tradable asset, computing will derive diverse business models such as compute trading, compute leasing, and compute sharing, forming a thriving compute economy ecosystem
  • Promoting inclusive sharing: Assetization lowers the barriers and costs of compute access, enabling small and medium institutions and individual developers to obtain needed compute at reasonable cost
  • Advancing green development: Measurable compute assets make energy efficiency and carbon footprint trackable metrics, incentivizing compute providers to adopt green energy and energy-saving technologies

The cloud computing power framework proposed by UNBY is precisely guided by the concept of compute assetization, building a complete system from resource aggregation, intelligent scheduling, trusted metering, to value circulation, making computing truly a foundational digital asset that flows efficiently and is shared inclusively in the digital economy era.

5. System Architecture

5.1 Architecture Design Principles

UNBY's system architecture follows these design principles:

  • Layered decoupling: Each functional layer has clear responsibilities and standardized interfaces, enabling independent evolution and elastic scaling
  • Heterogeneous compatibility: Supporting unified access and management of diverse computing resources including CPUs, GPUs, and AI accelerators
  • Intelligent scheduling: Dynamic resource matching based on multi-dimensional metrics, achieving globally optimal allocation
  • Secure and trusted: End-to-end data security and verifiable metering mechanisms
  • Open standards: Following international cloud computing and distributed computing standards, ensuring interoperability

5.2 Four-Layer Architecture Overview

UNBY adopts a four-layer architecture design, from bottom to top: Computing Resource Layer → Scheduling Engine Layer → Service Interface Layer → Application Ecosystem Layer. The layers work in concert to collectively support the full-process closed loop of compute from aggregation, scheduling, to consumption.

+---------------------------------------------+
|        Application Ecosystem Layer (4)      |
|   AI Training . Science . 3D Render . Edge  |
+---------------------------------------------+
|        Service Interface Layer (3)           |
|   SDK . API . Developer Toolchain           |
+---------------------------------------------+
|        Scheduling Engine Layer (2)           |
|   Smart Scheduling . Metering . Monitoring  |
+---------------------------------------------+
|        Computing Resource Layer (1)          |
|   Data Centers . GPU Clusters . Edge Devices|
+---------------------------------------------+

5.3 Layer One: Computing Resource Layer

The computing resource layer is the physical foundation of the entire network, responsible for aggregating and managing all types of heterogeneous computing resources.

Resource Types:

  • Data center compute: High-performance computing resources from hyperscale and mid-size data centers
  • GPU clusters: High-density GPU computing resources for AI training and inference
  • Edge devices: Lightweight computing nodes deployed at the network edge, suitable for low-latency scenarios
  • Personal idle compute: Idle computing capacity from personal computers and workstations

Core Functions:

  • Resource registration and discovery: Compute providers register resource information including compute type, specifications, geographic location, and availability windows
  • Health monitoring: Real-time monitoring of resource node operational status, load levels, and availability
  • Heterogeneous adaptation: Standardized adaptation layer to mask hardware architecture differences, enabling unified management
  • Security isolation: Ensuring resource isolation and data security between different users' tasks

The design goal of the computing resource layer is to build an open-access, heterogeneous-compatible global computing resource pool, maximizing the scale and diversity of schedulable compute.

5.4 Layer Two: Scheduling Engine Layer

The scheduling engine layer is the "brain" of the UNBY network, responsible for intelligent matching and global optimization of computing resources.

Core Functions:

Intelligent Scheduling

Precise matching of compute supply and demand based on multi-dimensional metrics. Scheduling decisions comprehensively consider the following factors:

  • Task characteristics: Compute type (training/inference/rendering), specification requirements, estimated duration
  • Resource status: Available compute, current load, network latency, geographic location
  • Cost efficiency: Resource utilization efficiency, energy consumption level, scheduling distance
  • Quality of service: Historical reliability score, task completion rate

The scheduling engine uses intelligent algorithms to find the optimal resource matching scheme globally, maximizing resource utilization efficiency and overall benefit while meeting task requirements.

Trusted Metering

Establishing a standardized compute measurement system for precise metering of compute contributions and consumption:

  • Using multi-dimensional metrics including FLOPS, throughput, and latency for comprehensive assessment
  • The entire metering process is traceable and verifiable
  • Ensuring that compute providers' contributions and compute consumers' usage are fairly recorded

Monitoring & Operations

  • Real-time monitoring and visualization of network-wide resource status
  • Task execution tracking and anomaly alerting
  • Automatic fault detection and failover
  • Network performance analysis and optimization recommendations

5.5 Layer Three: Service Interface Layer

The service interface layer is the bridge connecting the computing network with external applications, providing standardized access methods for developers and users.

Core Components:

  • SDK (Software Development Kit): Multi-language development libraries, supporting developers in quickly integrating compute scheduling capabilities
  • API (Application Programming Interface): Standard interfaces supporting compute querying, task submission, status tracking, and other operations
  • Developer toolchain: Including task orchestration tools, debugging tools, performance analysis tools, covering the full development lifecycle
  • Identity & access management: Unified identity authentication and access control system, ensuring resource access security

The service interface layer design emphasizes ease of use and openness, lowering the technical barriers for developers to access the computing network and promoting ecosystem prosperity.

5.6 Layer Four: Application Ecosystem Layer

The application ecosystem layer is the ultimate manifestation of the computing network's value, carrying various compute consumption scenarios and applications.

Core Application Domains:

  • AI model training & inference: Large language models, multimodal models, recommendation systems, etc.
  • Scientific computing: Genome sequencing, drug discovery, climate modeling, astrophysics, etc.
  • 3D rendering: Film VFX, architectural visualization, game development, etc.
  • Edge computing: Smart cities, autonomous driving, industrial IoT, etc.

The application ecosystem layer attracts third-party developers through an open platform to build diverse applications, forming rich compute consumption scenarios that in turn drive continuous expansion of the computing resource layer.

5.7 Architecture Coordination Mechanism

The four layers do not operate in isolation but form a closed loop through standardized interfaces and coordination mechanisms:

  • Top-down demand transmission: The application ecosystem layer generates compute demand, transmitted through the service interface layer to the scheduling engine layer, which matches optimal resources in the computing resource layer
  • Bottom-up supply feedback: The computing resource layer reports resource status in real time, the scheduling engine dynamically adjusts scheduling strategies, and the service interface layer feeds results back to the application layer
  • End-to-end observability: From resource registration, task scheduling to result delivery, the full-process data is traceable and auditable

This layered, decoupled, coordinated closed-loop architecture design enables UNBY to achieve efficient scheduling and circulation of computing resources while ensuring system flexibility and scalability.

6. Ecosystem

6.1 Ecosystem Overview

The UNBY ecosystem is a multi-party, value-co-creating open network. Five core participant types each play their roles and are interdependent, collectively driving the efficient flow of computing resources and value creation, forming a self-reinforcing ecosystem flywheel.

6.2 Five Core Participant Types

6.2.1 Compute Providers

Compute providers are the supply-side foundation of the network, contributing computing resources to serve network-wide demand.

  • Data center operators: Providing large-scale, highly reliable professional computing resources
  • GPU cluster owners: Providing high-performance GPU compute for AI computing
  • Edge device operators: Providing lightweight computing deployed at the network edge
  • Personal compute contributors: Contributing personal device idle compute, participating in resource pool building

Compute providers receive value returns proportional to their contributions, with returns correlated to the amount, quality, and usage duration of contributed compute.

6.2.2 Compute Consumers

Compute consumers are the demand-side driver of the network, using computing resources to complete various computational tasks.

  • AI enterprises: Conducting large model training, inference deployment, algorithm optimization
  • Research institutions: Conducting genome sequencing, drug discovery, climate modeling, and other scientific research
  • Creative industries: Executing 3D rendering, film VFX, game development, and other creative tasks
  • IoT application providers: Running smart city, autonomous driving, industrial IoT, and other edge computing applications

6.2.3 Compute Platform

The compute platform is the hub connecting supply and demand, providing scheduling, metering, settlement, and operations services.

  • Scheduling service: Intelligent matching of compute supply and demand
  • Metering service: Standardized measurement of compute contributions and consumption
  • Settlement service: Fair and transparent value settlement
  • Operations service: Ensuring network stability and resource health

6.2.4 Developer Community

The developer community is the core engine of ecosystem innovation, building applications and tools based on platform capabilities.

  • Application developers: Developing compute applications for end users
  • Tool developers: Building SDKs, API extensions, and development toolchains
  • Algorithm engineers: Optimizing scheduling algorithms and metering models

The developer community accesses the network through SDKs, APIs, and toolchains, contributing wisdom and creativity to enrich the ecosystem's application and tool system.

6.2.5 Governance Participants

Governance participants are responsible for maintaining the fairness, transparency, and sustainable development of the ecosystem.

  • Governance committee: A decision-making body composed of multi-party representatives
  • Community members: Participating in proposal voting and ecosystem building
  • Audit organizations: Independent third parties conducting transparency audits

6.3 Value Flow Mechanism

Value flow in the UNBY ecosystem follows a "contribution — metering — return" closed-loop mechanism:

Compute Providers --contribute compute--> Compute Platform --schedule & allocate--> Compute Consumers
     ^                                          |
     |              Value Return                 |
     +----------- Settle by contribution <-- use compute -+
  • Compute providers register resources to the platform, contributing available compute
  • Compute platform intelligently schedules based on consumer demand, allocating compute to corresponding tasks
  • Compute consumers use compute to complete tasks, generating usage records
  • Platform metering system precisely records each party's contributions and consumption
  • Settlement mechanism distributes value returns to compute providers based on contribution
  • Developers receive ecosystem incentives by building applications and tools

This closed loop ensures that every compute contribution receives fair returns, and every compute use is reliably guaranteed, driving the ecosystem's virtuous operation.

6.4 Ecosystem Flywheel Effect

The UNBY ecosystem has a self-reinforcing flywheel effect:

  • More compute providers join → Resource pool expands, supply diversity improves
  • Resource pool expands → Consumers access better, lower-cost compute → More consumers join
  • More consumer demand → Compute utilization rises, provider returns increase → Attracts more providers
  • Supply and demand both flourish → Platform scale effects emerge → Attracts more developers to build applications
  • Applications proliferate → Compute consumption scenarios expand → Demand further grows
  • Ecosystem prospers → Governance participants increase → Ecosystem becomes healthier and more sustainable

This flywheel effect gives the UNBY ecosystem powerful network effects and scale effects: the more participants, the greater the ecosystem value; the greater the value, the more it attracts new participants. As the ecosystem continues to grow, UNBY will gradually establish a first-mover advantage and network barriers in the compute economy domain.

7. Use Cases

The UNBY cloud computing network serves four core application domains, providing efficient, elastic, and economical compute solutions for different types of computational demands.

7.1 AI Model Training & Inference

Demand Analysis

Artificial intelligence is currently the fastest-growing area of compute demand. Training and inference for large language models (LLMs), multimodal models, recommendation systems, and other AI applications place extremely high demands on compute:

AI large model training compute demand is growing exponentially, doubling every 3-4 months.
  • Training phase: Requires massive GPU compute for large-scale parallel computation, with long training cycles and high resource consumption
  • Inference phase: Requires stable and reliable compute to support model deployment and online services, sensitive to latency
  • Iteration frequency: AI models iterate rapidly, with compute demand being continuous and frequent

In the traditional model, accessing large-scale AI training compute faces high costs, long queue times, and insufficient flexibility — particularly forming a significant barrier for small and medium AI teams.

Solution

UNBY provides elastic, on-demand compute supply for AI computing:

  • Elastic training clusters: Dynamically allocating GPU compute based on training task scale, without needing long-term fixed resources
  • Heterogeneous compute coordination: Integrating different specifications of GPUs and AI accelerators, optimizing training efficiency and cost
  • Training-inference integration: Smoothly transitioning from training to inference deployment after training, with seamless resource utilization
  • Distributed training support: Cross-regional compute resource coordination, supporting large-scale distributed model training

Value Creation

  • Lowering compute access costs for AI teams, accelerating model R&D iteration
  • Enabling small and medium AI teams to access high-quality training compute, promoting AI innovation democratization
  • Improving overall GPU resource utilization, reducing compute idle waste

7.2 Scientific Computing

Demand Analysis

Scientific research has a long-standing and vigorous demand for high-performance computing:

  • Genome sequencing: Massive genomic data processing and analysis, compute-intensive tasks
  • Drug discovery: Molecular simulation, virtual screening, protein structure prediction, requiring large-scale parallel computing
  • Climate modeling: Global climate model computation, with massive data volumes and high computational precision requirements
  • Astrophysics: Universe simulation, gravitational wave data analysis, and other frontier research

Research institutions generally face challenges of scarce supercomputing resources, long application cycles, and high usage costs, with many important research projects progressing slowly due to compute bottlenecks.

Solution

UNBY provides high-performance, scalable compute support for scientific computing:

  • High-performance computing clusters: Integrating data-center-level compute resources, supporting large-scale scientific computing
  • On-demand scaling: Elastically expanding compute scale based on research project computational demands
  • Data-proximity computing: Combining resource geographic location, optimizing data transfer and computing efficiency
  • Multi-precision computing support: Supporting scientific computing tasks with different precision requirements

Value Creation

  • Shortening scientific computing wait times, accelerating the pace of scientific discovery
  • Lowering compute access barriers and costs for research institutions
  • Enabling more research teams to conduct compute-intensive frontier research

7.3 3D Rendering

Demand Analysis

3D rendering is a core compute demand of the creative industry, covering film VFX, architectural visualization, game development, and other domains:

  • Film VFX: High-precision scene rendering, with single-frame rendering taking hours, demanding enormous compute
  • Architectural visualization: Architectural renderings and walkthrough animations, with large compute demand fluctuations during project cycles
  • Game development: Light baking, material rendering, requiring large batches of parallel rendering nodes

Rendering tasks are characterized by high burstiness, high parallelism, and good divisibility, making them well-suited for distributed computing network scheduling. Traditional render farms have high construction costs and volatile utilization, making them difficult for small creative teams to afford.

Solution

UNBY provides flexible, economical distributed rendering compute for 3D rendering:

  • Distributed render farm: Decomposing rendering tasks across multiple compute nodes for parallel execution, dramatically shortening render time
  • Elastic scaling: Flexibly adjusting the number of rendering nodes based on project progress, avoiding resource idle
  • Idle compute utilization: Integrating personal workstations and other idle compute to participate in rendering, lowering cost
  • Render task management: Providing task queuing, priority management, result return, and other complete functionality

Value Creation

  • Dramatically lowering rendering costs and wait times for creative teams
  • Enabling small studios to access render-farm-level compute support
  • Improving the utilization of globally idle graphics compute

7.4 Edge Computing

Demand Analysis

Edge computing is becoming an important complement to cloud computing, playing a key role in low-latency, data privacy, and bandwidth optimization scenarios:

The edge computing market is expanding rapidly, expected to reach hundreds of billions of dollars by 2027.
  • Smart cities: Video analysis, traffic management, environmental monitoring, requiring edge real-time processing
  • Autonomous driving: Vehicle perception and decision-making require millisecond-level low-latency computing
  • Industrial IoT: Equipment condition monitoring, predictive maintenance, quality control, and other real-time applications

Edge computing scenarios are characterized by geographically distributed deployment, latency sensitivity, and strong data locality, requiring computing resources deployed near the data source.

Solution

UNBY's network integrates global edge computing resources, providing proximity compute services for edge computing scenarios:

  • Edge compute discovery: Automatically matching the nearest edge compute node based on application location
  • Low-latency scheduling: Prioritizing geographically proximate compute resources, meeting latency requirements
  • Cloud-edge coordination: Edge compute and central compute coordinating, with layered processing of heavy and light tasks
  • Edge application deployment: Supporting rapid deployment and dynamic migration of edge applications

Value Creation

  • Providing low-latency compute guarantees for smart city, autonomous driving, and other scenarios
  • Lowering the cost and complexity of edge application compute deployment
  • Activating globally distributed edge computing resources, forming a wide-coverage compute network

7.5 Scenario Synergy & Value Resonance

The four application scenarios do not exist in isolation but share the UNBY compute resource pool, forming synergistic effects:

  • Resource complementarity: AI training primarily uses GPU clusters, scientific computing mainly uses high-performance CPUs, rendering needs graphics compute, and edge needs lightweight nodes — the differentiated demand for compute types across scenarios ensures all types in the resource pool are fully utilized
  • Time-window complementarity: Different scenarios have different peak demand periods, and cross-scenario scheduling can smooth resource load fluctuations
  • Technology spillover: Technological innovations from one scenario can benefit others — for example, distributed training technology from AI can be applied to scientific computing

This multi-scenario synergy makes the overall resource utilization of the UNBY computing network far exceed that of single-scenario applications, achieving maximum compute value.

8. Governance Mechanism

8.1 Governance Philosophy

UNBY upholds a governance philosophy of openness, transparency, and co-governance. As a multi-party infrastructure, the healthy development of the computing network depends on a fair and reasonable governance mechanism. UNBY is committed to building a community-driven, multi-party balanced, continuously evolving governance system, ensuring the network consistently serves the overall interests of the ecosystem over the long term.

8.2 Community Governance Structure

UNBY's governance structure adopts a multi-layered design:

Governance Committee

The governance committee is the core decision-making body of the ecosystem, composed of multi-party representatives:

  • Compute provider representatives: Representing supply-side interests
  • Compute consumer representatives: Representing demand-side interests
  • Developer representatives: Representing the technical community's interests
  • Independent scholars/industry experts: Providing neutral professional perspectives
  • Ecosystem operator representatives: Representing the platform operator

The governance committee is responsible for reviewing major decisions, including network upgrades, parameter adjustments, and rule revisions, ensuring all parties' interests are balanced.

Community Assembly

The community assembly is a broad-based mechanism for expressing public opinion, held regularly and open to all ecosystem participants. Community members can propose suggestions, express concerns, and participate in discussions at the assembly, forming community consensus.

Specialized Working Groups

Specialized working groups are established for specific domains, such as the technical standards group, security audit group, and green development group, responsible for in-depth research on specific issues and submitting recommendations to the governance committee.

8.3 Proposal Process

UNBY adopts a standardized community proposal system, ensuring democratic and transparent governance decisions:

  • Proposal initiation: Any eligible community member can initiate a proposal, which must include problem description, solution, and impact analysis
  • Community discussion: Proposals enter a public discussion period, where community members exchange views and suggest modifications
  • Technical assessment: The technical working group conducts a professional assessment of the proposal's feasibility and impact
  • Voting: The governance committee and/or community votes, with the outcome determined by voting rules
  • Implementation: Approved proposals enter the execution phase, with the process publicly trackable
  • Effect review: After implementation for a period, an effectiveness evaluation is conducted, and a revision process may be initiated if necessary
The community proposal system ensures every participant's voice is heard, guaranteeing that governance decisions reflect the collective will of the ecosystem.

8.4 Transparency Assurance

Transparency is the cornerstone of governance credibility. UNBY ensures governance transparency through the following mechanisms:

  • Transparency reports: Regularly publishing network operations transparency reports, disclosing key data such as resource scale, scheduling efficiency, and settlement fairness
  • Open decision-making: The governance committee's review processes and decision outcomes are made public to the community
  • Auditable data: Compute metering, scheduling records, and settlement data are fully traceable, supporting independent audits
  • Public information portal: Establishing a unified information disclosure platform for community access to governance-related information

8.5 Security Audit

Security is the lifeline of the computing network. UNBY establishes a multi-layered security assurance system:

Technical Security

  • Resource isolation: Strict isolation of computing resources and data between different users' tasks
  • Data protection: End-to-end data security with transmission encryption, storage encryption, and access control
  • Identity authentication: Unified identity authentication and permission management system
  • Tamper-proof mechanisms: Metering and settlement data are tamper-proof, ensuring record credibility

Independent Audit

  • Regular security audits: Engaging independent third-party security organizations for periodic system security audits
  • Code audits: Core system code is subject to open-source community and third-party audits
  • Operational audits: Continuous monitoring and independent auditing of network operational status

Incident Response

  • Vulnerability response: Establishing vulnerability reporting and response mechanisms, promptly fixing security issues
  • Emergency plans: Developing comprehensive emergency response plans, ensuring network stability under abnormal conditions
  • Incident notification: Timely notification to the community after security incidents, maintaining information transparency

Through comprehensive governance mechanisms, transparency assurance, and security audit systems, UNBY is committed to building a computing network infrastructure worthy of all participants' trust.

9. Development Roadmap

9.1 Roadmap Overview

UNBY's development plan is divided into four phases, following a progressive path of "build the foundation — expand the ecosystem — global deployment — intelligent upgrade." Each phase sets clear objectives, with the results of the preceding phase laying the foundation for the next.

9.2 Phase One: Infrastructure Construction

Phase Objective: Complete core system build, achieving basic compute scheduling capability.

Key Tasks:

  • Complete the design and development of the four-layer architecture core modules
  • Establish compute resource registration, discovery, and basic scheduling capabilities
  • Implement the standardized compute measurement system
  • Onboard the first batch of data center and GPU cluster computing resources
  • Complete SDK and basic API development, supporting developer access
  • Establish basic security protection and monitoring systems

Phase Outcome: Form a runnable minimal computing network prototype, validating the feasibility of core scheduling and metering mechanisms.

9.3 Phase Two: Ecosystem Expansion

Phase Objective: Expand resource scale and application scenarios, establishing an initial ecosystem flywheel.

Key Tasks:

  • Expand compute resource access types, supporting edge devices and personal idle compute
  • Enhance the scheduling engine to support multi-dimensional intelligent matching and cross-regional scheduling
  • Launch developer toolchain, lowering ecosystem access barriers
  • Onboard AI training, scientific computing, and other core application scenarios
  • Establish community governance framework, launch community proposal mechanism
  • Publish the first transparency report

Phase Outcome: Compute resource pool scale significantly expanded, multi-scenario applications deployed, and ecosystem participants forming initial virtuous interactions.

9.4 Phase Three: Global Deployment

Phase Objective: Achieve global-scale compute resource integration and cross-regional scheduling.

Key Tasks:

  • Advance multi-region compute node deployment globally
  • Implement cross-regional, cross-time-zone intelligent compute scheduling
  • Establish international compliance frameworks, adapting to regulatory requirements across regions
  • Expand international compute provider and consumer bases
  • Enhance governance mechanisms, introducing international governance participants
  • Deepen green computing practices, prioritizing integration of renewable-energy compute

Phase Outcome: Build a distributed computing network covering major global regions, achieving efficient cross-regional compute flow.

9.5 Phase Four: Intelligent Upgrade

Phase Objective: Introduce intelligent technologies, achieving autonomous optimization and continuous evolution of the computing network.

Key Tasks:

  • Introduce intelligent algorithms to optimize scheduling decisions, achieving adaptive resource matching
  • Build compute demand prediction models, enabling resource pre-scheduling and forward-looking resource allocation
  • Refine the compute assetization system, expanding compute circulation and trading scenarios
  • Deepen the ecosystem, supporting more innovative application forms
  • Establish a comprehensive ecosystem standards system, driving industry standard development
  • Continuously optimize energy efficiency, building a green, low-carbon computing network benchmark

Phase Outcome: UNBY becomes an intelligent global computing network with autonomous optimization capabilities, with a mature and refined compute assetization ecosystem.

9.6 Phase Transitions & Continuous Evolution

The four phases are not rigid time divisions but an interconnected, iteratively advancing process. UNBY will dynamically adjust the pace and priorities of each phase based on technology development, market demand, and community feedback. At the end of each phase, a comprehensive evaluation is conducted, summarizing experience, identifying shortcomings, and guiding the next phase.

UNBY always adheres to long-termism, advancing computing network construction in a sustainable manner, ensuring every step of development withstands the test of time.

10. Team & Advisors

10.1 Core Team

UNBY was founded by a cross-disciplinary, cross-domain professional team, with members possessing deep expertise and practical experience in cloud computing, distributed systems, artificial intelligence, and computing infrastructure.

Technical Team

  • Core technical personnel have experience in large-scale distributed system design and operations
  • Deep research in cloud computing resource scheduling and heterogeneous computing architecture optimization
  • Have participated in the construction and operation of large data centers and computing platforms

Product & Ecosystem Team

  • Experience in cloud computing product planning and developer ecosystem operations
  • Deep understanding of compute market supply-demand characteristics and user pain points
  • Committed to building an easy-to-use, open compute service platform

Operations & Governance Team

  • Experience in community governance and ecosystem operations
  • Familiarity with compute industry policies, regulations, and compliance requirements
  • Driving transparent, fair governance mechanism development

[Specific team information to be added]

10.2 Advisory Team

UNBY brings together senior advisors from academia and industry, providing strategic guidance for project development:

  • Academic advisors: Renowned scholars in cloud computing, distributed computing, and computing networks, providing frontier technology direction guidance
  • Industry advisors: Senior practitioners from data center, AI, and cloud computing industries, providing market insights and industry resource connections
  • Compliance advisors: Professionals familiar with laws and regulations related to digital assets and compute services, ensuring compliant project operations
  • Sustainable development advisors: Experts with rich experience in green computing and energy efficiency, guiding green compute practices

[Specific advisor information to be added]

10.3 Team Culture

The UNBY team upholds the following core values:

  • Technology conviction: Believing in the power of technology to change compute resource allocation
  • Open collaboration: Building the compute ecosystem with an open mindset alongside the community
  • Long-termism: Guided by sustainable development, not pursuing short-term gains
  • User first: Always designing products and services centered on compute users' needs

The team will continue to attract like-minded outstanding talent to jointly advance the realization of the UNBY vision.