Project Overview · 2026

Solid Waste Industry Ecological AI Large-Model Intelligent Computing Center

固废产业生态AI大模型智算中心 · 专业固废资源化产业服务平台

A professional service platform for the solid waste recycling industry: an industry-specific AI large model trained on real project data, backed by a 1,000 PFLOPS green computing center powered by 100% renewable energy.

CIWM · Solid Waste Industry Research Center Huazhu Green Building (Beijing) Technology Co., Ltd. 1,000 PFLOPS · PUE ≤ 1.20 · Zero-carbon computing

Executive Summary

The Challenge

China's cumulative stockpile of industrial solid waste exceeds 33 billion tonnes, with about 4.5 billion tonnes generated each year. The new Mineral Resources Law makes green mine construction a statutory obligation, and the national 15th Five-Year Plan will invest over RMB 26 trillion in the "Six Networks" programme — a powerful policy tailwind for the industry.

The Solution

The Solid Waste Industry Ecological AI Large-Model Platform acts as the "intelligent brain" of mining, processing and metallurgy value chains, enabling "zero-waste" upgrading. A dedicated green intelligent computing center provides the compute backbone, powered by 100% renewable energy for a low-carbon, high-efficiency infrastructure.

The Advantage

Built on real data from CIWM's industrialized projects, the platform creates a technology moat that is hard to replicate and trains a world-leading industry AI model. Revenue is diversified: near-term project replication service fees recover capital quickly, while long-term income comes from computing leasing, SaaS subscriptions and tiered industry services.

Background & Policy Opportunity

Industry Pain Points

  • Stockpiles & ecological pressure — unregulated dumps occupy land and pollute soil and water.
  • Wasted resources — conventional processes recover little; valuable minerals are discarded with the slag.
  • Experience-driven, ageing workforce — no standardized knowledge system; experts are scarce.
  • Data silos — no unified industry data platform; information cannot be shared or used for intelligent decisions.

Policy & Market Tailwinds

  • New Mineral Resources Law (in force July 2025): green mine construction upgraded from policy encouragement to statutory duty.
  • Hard targets by end-2028: ≥90% compliance for large mines, ≥80% for medium mines.
  • Computing-infrastructure dividend: over RMB 4 trillion of direct new investment in the computing network alone.
  • Green-power advantage: 50.9% green electricity direct supply in Inner Mongolia's computing hub and Baotou's national pilot zone.
US$10.87 BnGlobal mining-industry large-model market (2026 forecast)
US$68 BnGlobal mine solid waste comprehensive utilization market (2026)
RMB 750 BnChina industrial solid waste utilization output value

Sector CAGR for 2025–2030 is expected to reach 55–70% as applications move from pilots to scale.

Positioning & Goals

Vertical Industry Model

Deeply trained on real project data — purpose-built for the sector, not a generalized model. Strong industry specificity and practical applicability.

AI-Agent Decision Making

Capable of autonomous reasoning: beyond analytics, it delivers complete solutions from scheme design to on-site implementation.

Full Lifecycle Coverage

From source reduction and process control to end-of-pipe treatment — green enablement across the entire chain.

Platform Architecture — Four Layers

Algorithm Layer · Top-tier Compute Intelligence

Integrates core algorithms from 10 leading global AI companies, reusing tens of billions of RMB in R&D results as a solid foundation.

Model Layer · Dual-Brain Architecture

An innovative "cognitive brain + executive brain" model links strategic decision-making with micro-level execution control for high-precision decisions.

Data Layer · Nationwide Resource Base

A national solid waste resource database covering distribution, categories and industrial analysis — breaking data silos.

Application Layer · Multi-Scenario Services

Customized intelligent services for governments, enterprises and investors: decision support, operations optimization and value assessment.

Goal: a nationally leading AI computing hub for solid waste recycling, driving the deep integration of green, low-carbon and efficient industry development.

Application Scenarios

Integrated Project Solutions

End-to-end technical and management solutions covering feasibility, design and operations of recycling projects.

Serves: solid waste disposal enterprises

Mine Ecological Restoration Planning

AI-assisted, site-specific restoration plans balancing ecological recovery, land reclamation and biodiversity.

Serves: mining companies, local governments

Zero-Waste Development Planning

Customized "zero-waste mining-processing-metallurgy" schemes that optimize resource pathways and cut waste at source.

Serves: mining operators

Investment Decision Support

Big-data and model-based assessment of project economics and environmental risk for scientific investment.

Serves: financial institutions, investors

Smart Mining & Beneficiation Optimization

AI-optimized production raising ore recovery and separation precision for efficient enrichment of valuable minerals.

Serves: mining and mineral processing companies

Policy Compliance & Supervision Support

Automated green-mine compliance assessment, carbon accounting and environmental monitoring for smart regulation.

Serves: environmental and mining regulators

Zero-Waste Smelting Solutions

Cleaner production systems for smelters that solve slag disposal and enable green circular smelting.

Serves: non-ferrous metal smelters

Full-Chain Enablement

Covering the whole lifecycle from mine development and processing to restoration, investment and supervision — a win-win for all stakeholders.

The Green Computing Center

1,000 PFLOPSTotal compute (FP16/BF16) for large-scale AI training and HPC
80–110 racksStandard racks, 22–32 kW per rack, high-density ready
PUE ≤ 1.20Meeting mandatory national hub-node standards
Cold-plate liquid coolingPrimary cooling with hybrid air-assist for efficient heat dissipation
100% renewableFull waste-heat recovery — a zero-carbon computing benchmark
RMB 0.21/kWhGreen power direct-connect net price, far below eastern China

Site Advantages

  • Policy fit — aligned with the national "Six Networks" computing and energy programmes.
  • Abundant green power — Inner Mongolia's installed renewable capacity reaches 180 GW.
  • National pilot status — Baotou is a national new-type power-system pilot with clear policy support.
  • Industrial cluster synergy — rich local demonstration projects accelerate data collection and model training.

Green & Low-Carbon Operation

Advanced liquid-cooling architecture achieves efficient heat dissipation for high-density compute, sharply cutting energy use — a core infrastructure for the "East Data, West Computing" strategy and green, low-carbon operations.

Flagship Demonstration Projects

Four existing CIWM projects provide real-world cases and templates for the AI model's development and application, covering non-metallic mines, polymetallic mines and hazardous waste treatment.

Shangluo Non-Metallic Mine

A zero-waste mining and beneficiation demonstration combining green extraction with efficient resource utilization.

Pingshan Polymetallic Mine

A polymetallic zero-waste mining and processing system for refined utilization in harmony with the environment.

Guyang Zhongsheng Environmental (Aluminum)

Pioneering zero-waste smelting for the electrolytic aluminum industry and circular-economy hazardous waste treatment.

Inner Mongolia Kangdeli Environmental (Batteries)

Efficient recovery of precious metals and a full-lifecycle green closed loop for spent batteries.

From demonstration to standards: converting proven project experience into industry standards leads the direction of green mines and zero-waste plants — and builds a durable technology moat.

Business Model

Data Services — RMB 1,000

Real-time solid waste data queries and industry updates.

Business Plans — RMB 5,000

AI-assisted, customized project development business plans.

Feasibility Studies — RMB 10,000

Complete professional documentation for project filing and approval.

Financing Matchmaking — RMB 100,000+

Customized financing schemes with direct access to quality investors.

Integrated Solutions — from RMB 1 Mn

Full-process delivery from development planning and construction to operation.

Revenue Mix at Full Capacity

Project services 50% · Computing leasing 20% · Data services 20% · Consulting 10%.

Investment & Financials

RMB 1.18 BnTotal investment — equipment 61% · construction & liquid cooling 26% · other & contingency 13%
30 / 20 / 50Funding mix — equity capital / government industry funds & special bonds / syndicated loans (RMB 350 / 240 / 590 million)
RMB 400 MnAnnual revenue at full capacity across four service lines
RMB 320 MnGross profit at full capacity — high-margin profile
3–4 yearsStatic payback period — fast capital recovery
15–22%After-tax IRR on total investment — well above industry benchmarks

Implementation Roadmap

Phase 1 · Approval & Feasibility

Months 1–6: project filing, energy review, EIA and grid-connection approvals.

Phase 2 · Civil & M&E Works

Months 7–18: building construction, power distribution, precision liquid cooling and fire/safety systems.

Phase 3 · Compute Deployment

Months 12–18: AI server clusters, high-speed switching and distributed storage installation and commissioning.

Phase 4 · Platform & Model Development

Months 6–24: compute scheduling platform, industry AI model training and national data foundation.

Phase 5 · Trial Run & Acceptance

Months 19–24: full-system stress testing, security accreditation and completion acceptance.

Phase 6 · Scale Operations

Month 25 onwards: full-capacity operation delivering intelligent "zero-waste" solutions for green development.

Conclusion

Standing on major national policy dividends and Inner Mongolia's green computing hub advantages, the project uses a 1,000 PFLOPS green intelligent computing center as its technical base to build a world-leading vertical AI model for solid waste recycling — delivering intelligent solutions for the industry's green transformation.

Strong Policy Alignment

Deeply aligned with "East Data, West Computing", green mine construction and the "Six Networks" national strategies.

Significant Technology Moat

Trained on real full-process industrial data, with proprietary zero-waste implementation technology that is hard to replicate.

Clear Business Model

"Computing services + technology licensing + operations revenue sharing" — high-margin, stable cash flows.

Attractive Returns

Payback in about three years with an IRR well above industry averages — strong investment value and growth potential.

This English page is a reference translation prepared for international visitors. The official Chinese presentation prevails: view the Chinese original →