Changpeng Wan
– The AI computing sector is demonstrating incredible money-making potential, with major ETFs seeing gains of over 46% since April.
– Massive capital expenditures from cloud giants—projected at $359 billion for 2025—are fueling unprecedented growth in AI infrastructure.
– Chinese optical module manufacturers dominate the global market with over 60% share, led by companies like InnoLight and Eoptolink.
– The transition to 1.6T optical modules presents the next major growth frontier, with ASPs expected to be 50-100% higher than current 800G modules.
– Despite recent gains, valuations remain reasonable with the AI revolution still in its early innings, creating ongoing investment opportunities.
The landscape of global technology investing has been fundamentally reshaped by the artificial intelligence revolution, creating what many analysts are calling an incredible money-making potential in specific market segments. While broader indices experienced volatility after breaking through the 3,700-point threshold, AI-related sectors notably defied the trend, with CPO (Co-Packaged Optics) and artificial intelligence segments emerging as leading gainers. This divergence highlights how specific technology niches are creating extraordinary wealth generation opportunities that transcend broader market movements.
Companies like VeriSilicon Holdings Co., Ltd. saw gains exceeding 13%, while industry leaders Eoptolink Technology Inc. Ltd. and InnoLight Technology Corporation both surged over 5%. These moves propelled specialized ETFs like the Science and Technology Innovation Board AI ETF (588730) and the Artificial Intelligence ETF (159819) significantly higher. This impressive performance isn’t just a short-term phenomenon—since April 9th, the Artificial Intelligence ETF (159819) has skyrocketed more than 46%, while the Science and Technology Innovation Board AI ETF (588730) has gained over 36%. Even with these substantial advances, the underlying momentum behind the AI revolution suggests this incredible money-making potential remains largely intact.
What’s Driving This Extraordinary Performance?
The sustained strength in CPO concept stocks stems from multiple converging factors, with several AI-specific developments providing immediate catalysts. Shanghai’s recently released “AI+Manufacturing” implementation plan outlines ambitious goals to deepen integration between artificial intelligence technologies and manufacturing processes, accelerating the empowerment of new industrialization and forming new quality productive forces. Simultaneously, the upcoming 2025 China Computing Power Conference in Datong, Shanxi, scheduled for August 22-24, will showcase the latest achievements across the computing ecosystem, including policy frameworks, core hardware, infrastructure, key technologies, and service capabilities.
Beyond these specific catalysts, robust fundamentals and a buoyant equity market provide the foundation for this incredible money-making potential. Interim financial reports offer compelling evidence: Eoptolink reported H1 2025 net profit growth exceeding 300% year-over-year, while InnoLight, though growing at a relatively slower pace, still achieved impressive growth exceeding 50%. The broader market context further supports this trend, with the Shanghai Composite Index recently touching 3,745 points—a near 10-year high—while total A-share market capitalization surpassed the monumental 100 trillion yuan threshold.
Valuation Perspective: Room for Growth?
Despite these advances, research from China International Capital Corporation Limited (CICC) indicates that comprehensive valuation metrics—including price-to-earnings ratios, market capitalization to GDP ratios, and dividend yields—suggest A-shares remain within reasonable ranges both historically and compared to global peers. The CSI 300 Index currently trades at approximately 12.2 times forward earnings, placing it around the 69th percentile historically since 2010. Compared globally, A-share valuations remain mid-range among major markets, with total market capitalization to GDP ratios positioned at medium-low levels among developed markets.
The ratio of total A-share market capitalization to M2 money supply stands at approximately 33%, at the 60th historical percentile. Meanwhile, the CSI 300’s dividend yield of 2.69% continues to offer relative attractiveness compared to ten-year government bond yields. Although daily trading volume has exceeded 2.8 trillion yuan with turnover rates surpassing 5% based on free-float market capitalization—historical patterns suggest such periods often experience increased short-term volatility without necessarily disrupting medium-term upward trajectories.
The Fundamental Drivers: Where Will Future Growth Originate?
The incredible money-making potential within the AI computing sector finds its strongest support in unprecedented capital expenditure commitments from cloud infrastructure giants. Recent H1 2025 financial reports revealed that North America’s four major cloud service providers have all maintained or increased their AI computing capital expenditure plans for the year. Microsoft leads with projected annual spending of $80 billion, followed by Meta with guidance of $66-72 billion, while Google has raised its 2025 capital expenditure to $85 billion (from originally planned $75 billion). Amazon anticipates exceeding $80 billion in capital expenditures for the year.
Collectively, these four behemoths are projected to deploy approximately $359 billion in 2025 capital expenditures—a staggering 57% year-over-year increase—with AI computing infrastructure representing the primary focus. While China’s computing infrastructure development has progressed somewhat slower, the catch-up effort is accelerating dramatically. The country’s three major telecom operators are increasingly tilting capital expenditures toward AI computing, with China Mobile planning 151.2 billion yuan in total 2025 capital expenditures, including 37.3 billion yuan dedicated specifically to computing infrastructure.
The Chinese Response to AI Infrastructure Demand
Alibaba publicly announced plans for several years of 380 billion yuan in capital expenditures, with AI infrastructure representing a central priority. Other technology leaders including Huawei, ByteDance, and Tencent are similarly accelerating AI computing infrastructure investments. Market research forecasts suggest China’s 2025 AI computing budget will reach approximately $120 billion (800 billion yuan), with roughly 50% allocated to AI chip procurement. The rationale is straightforward: without sufficient computing resources, competitiveness in the intensifying global AI race becomes unsustainable.
The computational demands of advanced AI models illustrate why this incredible money-making potential exists. OpenAI’s GPT-5 reportedly requires parameter scales approximately 10 times larger than GPT-4, with training computational requirements 4-8 times greater and inference computing needs 3-5 times higher. This explains why OpenAI CEO Sam Altman has consistently emphasized computational shortages, recently stating the company would invest trillions of dollars to solidify AI infrastructure supporting continuously growing computational demands for artificial intelligence services.
The Optical Module Opportunity: A Case Study in Value Creation
Within this expansive AI infrastructure ecosystem, optical module manufacturers represent a particularly compelling investment thesis demonstrating incredible money-making potential. The massive capital expenditures from global AI cloud service providers benefit the entire AI computing产业链, with optical module enterprises standing as clear beneficiaries. AI training clusters (such as thousand-GPU interconnections) have created surging demand for 800G optical modules, with 2025 global demand projected to exceed 18 million units—nearly doubling 2024 volumes.
The next technological transition is already underway, with 1.6T optical modules following NVIDIA’s GPU product upgrades (B200 series) beginning small-scale commercial deployment. Volume adoption is expected in 2026, but more significantly, 1.6T optical modules carry substantially higher average selling prices—projected at 50-100% premiums compared to 800G modules. This incredible money-making potential extends beyond immediate demand growth to include meaningful average selling price expansion.
Industry analysts project the optical module sector could maintain high overall performance (EPS) growth over the next three years, including doubling in 2025 and exceeding 50% growth in 2026. Chinese manufacturers particularly dominate this space, commanding over 60% of global market share, with leaders like InnoLight and Eoptolink maintaining continuous leadership. This competitive advantage positions them exceptionally well to capitalize on the global AI infrastructure buildout.
Valuation Expansion: Beyond Fundamentals
While fundamental growth provides the foundation, multiple factors support valuation expansion (PE multiple growth) that further enhances this incredible money-making potential. Improving geopolitical tensions, dissipating trade war uncertainties, and expected Federal Reserve rate cuts beginning in September all contribute to favorable macro conditions. Most significantly, strengthening equity market momentum—with U.S. tech stocks reaching new heights and NVIDIA and Microsoft repeatedly setting records—creates positive spillover effects for A-share chips, cloud computing, and especially optical module stocks.
A-shares returning to decade highs with total market capitalization breaking 100 trillion yuan attracts additional incremental capital, including growing global fund inflows. Morgan Stanley statistics indicate global long-only funds injected $2.7 billion into Chinese equity markets in July, accelerating from June’s $1.2 billion. International capital actively allocating to Chinese assets not only brings incremental funding but continuously enhances A-shares’ internationalization level, deepening connectivity between China’s quality enterprises and global markets while supporting economic recovery and industrial transformation—ultimately generating better returns for global investors.
Risk Appetite and Valuation Support
In this environment, risk appetite will likely continue ascending, potentially pushing market valuation benchmarks higher. Not only optical modules but other AI sub-sectors—including chips, cloud computing, AI applications, and even the broader technological innovation direction—benefit from these trends, becoming focal points for capital allocation. ETFs like the Science and Technology Innovation Board AI ETF (588730), which includes weightings from Cambricon Technologies Corporation Limited, Montage Technology Co., Ltd., Kingsoft Office Software, Inc., and VeriSilicon Holdings Co., Ltd., concentrate on the AI产业链’s most critical computing chip and intelligent hardware segments.
This ETF has attracted 137 million yuan over the past 20 trading days, with net inflows reaching 661 million yuan year-to-date and latest规模 reaching 1.003 billion yuan. The incredible money-making potential extends beyond individual stocks to encompass these thematic investment vehicles that provide diversified exposure to the AI infrastructure boom.
The Long-Term Investment Thesis for AI Infrastructure
Artificial intelligence is unquestionably leading a new technological revolution, accelerating penetration across countless industries and emerging as a core driver of global economic growth. As fundamental infrastructure supporting AI development, computing power faces unprecedented opportunities. On the demand side, computational requirements for training large AI models currently double every 3-4 months, with advanced models like GPT-5 requiring tens of thousands of high-end GPUs working协同—directly propelling rapid expansion of global AI chip markets.
Supply-side innovations equally contribute to this incredible money-making potential. Continuously upgraded GPUs, volume production of TSMC’s 3nm/2nm advanced processes significantly enhancing chip computational density, breakthroughs in Chiplet technology effectively alleviating memory bandwidth bottlenecks, proliferating liquid cooling solutions dramatically improving energy efficiency, and high-speed optical modules providing interconnection channels between data centers—all represent critical enabling technologies.
Navigating Current Valuations
After consecutive gains, valuations across optical modules, computing stocks, and other AI segments may not appear as attractive as during April’s market bottom. However, provided these companies maintain rapid growth momentum alongside strong market conditions, valuation digestion could occur relatively quickly. Medium to long-term, whether AI chips, optical modules, or mid-downstream AI segments including cloud services, software applications, and hardware applications—all will continuously benefit from AI industry development, with investment value likely to keep emerging.
The incredible money-making potential within AI infrastructure represents a long-term structural trend rather than short-term cyclical opportunity. Investors should focus on companies with sustainable competitive advantages, strong technological moats, and visible revenue growth trajectories rather than attempting to time short-term market fluctuations. The AI revolution remains in its early innings, suggesting this incredible money-making potential will likely continue unfolding across multiple years and technological upgrade cycles.
For investors seeking exposure to this transformative trend, specialized ETFs offering diversified access to the entire AI价值链 provide compelling options. The Artificial Intelligence ETF (159819), with its 0.15% management fee and 0.05% custody fee, offers particularly cost-efficient exposure to this incredible money-making potential. Meanwhile, direct investment in sector leaders like InnoLight and Eoptolink provides purer play on specific technology segments within the broader AI infrastructure theme.
As global economies increasingly digitize and intelligence becomes embedded across industries, the companies providing the fundamental computational infrastructure will likely continue demonstrating this incredible money-making potential that has captivated investors worldwide. The convergence of unprecedented technological demand, massive capital investment, and reasonable valuations creates a rare investment opportunity that transcends traditional market cycles and geographic boundaries.
