Evidence of China’s Computing Power Supply Strength

— Why the “China Supply + Singapore Clearing” Model is Globally Viable

I. Core Thesis

Our supply side does not need to be cultivated from scratch by us—it has already been validated by global capital with real financial commitments.

China has already produced a cluster of world-class AI compute and model providers priced as “trillion-dollar assets” by secondary markets. Our task is not to engineer supply, but to connect these existing world- class capacities to a tradeable, clearable, and exportable global market. AIEX Strategy Briefing | Confidential

II. Live Case Study: Zhipu AI (02513.HK) 指标 / Metric

Capital Market Status

“The World’s First Publicly Traded LLM Stock”

China’s model supply side has received the highest level of validation from global capital.

IPO Market Cap

57.9 billion HKD on day one (Jan 2026)

The baseline starting point is already a tens-of-billions USD supplier.

Latest Market Cap

Exceeded 1 Trillion HKD (2026.06.22)

Surged nearly 20x in half a year; capital markets are aggressively pricing “China’s cognitive capacity.”

Technical Validation GLM-4.6 / GLM-5.2 登顶

GLM-4.6 / GLM-5.2 topped CodeArena globally

The “goods” produced on the supply side are world-class quality, not low-end capacity.

Capacity Matrix

Full-stack: foundational, reflection, multimodal, autonomous agents

A complete matrix of token categories is available, supporting all-category trading.

Market Position

No. 1 in revenue among China’s independent general LLMs

Concentration of leading suppliers facilitates the exchange’s connection to standardized supply.

In short: A model company with 2024 revenue of just 312 million RMB and still running significant losses is valued by capital at 1 trillion HKD. This capital is not rewarding a single company; it is pricing “China’s cognitive capacity” as an entire asset class.

III. Extrapolation: Zhipu is Just the Tip of the Iceberg

Zhipu is not an isolated case; it represents China’s fully formed, world-class compute and model supply cluster: Model Layer: Leading model providers like Zhipu, DeepSeek, Tongyi, and Kimi continuously output world- class open-source and closed-source models. • • AIEX Strategy Briefing | Confidential

Compute Layer: China possesses the world’s strongest cost advantages + tidal off-peak dispatch capabilities (power, land, and hyper-scale data centers). Capacity Profile: 24/7/365 uninterrupted supply with continuously declining marginal costs — the exact ideal profile for a commoditized supply side. For an exchange, an ideal supply side must meet three criteria: massive volume, standardization, and low marginal cost. China’s supply side ticks all three boxes.

IV. The Closed Loop: Translating Supply Strength into Exchange Profit

World-Class Chinese Compute/Model Suppliers (e.g., Zhipu) → Output massive, low-cost, high-quality tokens ▼

Your AIEX (Singapore Neutral Clearing) → Standardization → Spot/Futures/Contracts/Bulk → Global Export ▼

Global Demand Side (Buyers restricted by compute regulations) → Secure trusted delivery & settlement in a neutral jurisdiction ▼

Platform Take Rate × $50 Trillion Notional Trading Volume

Core Logic: The stronger, more competitive, and lower- cost the supply side becomes, the more tradeable tokens flow into the exchange, driving up our matching and clearing volumes. While suppliers engage in intense margin-cutting competition at the production layer, we reap steady transaction fees at the circulation layer. • • • • AIEX Strategy Briefing | Confidential

V. Concluding Sentence for LPs

“Do not ask us where the supply side is—it is already listed on the HKEX with a market cap of one trillion.”

What we are doing is turning on the faucet to the global market for these world-class capacities stranded on one side of the geopolitical fault line. The value of the supply side has already been proven by global capital; our role is to prove the value of the liquidity and circulation layer.

Market capitalization data for Zhipu is sourced from public reports as of 2026/06/22. Stock prices fluctuate dynamically; please verify with real-time market data before finalization. It is recommended to denote “Source: Public Market Data as of [Date]” in the BP for institutional rigor. AIEX Strategy Briefing | Confidential