The Optical Mirage: Why AI Infrastructure Hype Will Dissolve Like DeFi Summer

Goldman Sachs just placed a staggering bet on Zhongji Xuchuang, the Chinese optical module manufacturer. By projecting profit growth of 65%, 108%, and 119% for 2026, 2027, and 2028, the investment bank has set a target price implying a 163.6% upside from current levels. The narrative is seductive: AI clusters demand faster interconnects, and Zhongji is the 1.6T/3.2T supplier for Nvidia. But as someone who spent 2019 manually tracing liquidity pools in Uniswap V1, I recognize the pattern. This is not fundamental strength—it is the same liquidity illusion that collapsed DeFi in 2022, now dressed in optical fiber. To understand the fragility, we must first map the global liquidity landscape. The current AI infrastructure boom is fueled by a flood of cheap capital from central banks and sovereign wealth funds, chasing a narrative of technological supremacy. Since 2023, global AI-related investments have surpassed $200 billion, with a significant portion directed at hardware—GPUs, networking, and optical modules. This is not organic demand; it is a policy-driven liquidity injection. The optical module supply chain—CW lasers, SOI silicon substrates, DSP chips—has become a leveraged bet on that liquidity continuing. But liquidity is a mirage; only settlement is real. The core insight from analyzing Zhongji Xuchuang’s technology is that its success hinges on engineering margins, not foundational innovation. The shift from 800G to 1.6T optical modules involves incremental improvements in electro-absorption modulators and silicon photonics integration. The technical challenges—signal integrity, thermal management, yield optimization—are real but solvable. What matters is the economic moat. During my DeFi Summer disillusionment in 2021, I realized that protocols without real-world utility amplify greed. Similarly, optical modules are a commodity; the surviving suppliers will be those with the best cost structure and customer lock-ins, not necessarily the most advanced technology. Goldman Sachs’ projection assumes Zhongji will maintain ASP premiums while competitors—newcos, Coherent, Huawei’s partners—fall behind. That assumption is dangerously optimistic. My contrarian angle is rooted in the decoupling thesis. Many analysts argue that AI hardware demand is decoupled from broader macroeconomic cycles. I disagree. The true decoupling is between hype and reality. In 2022, when AI capital expenditure growth slowed, the entire optical module sector saw inventory corrections. The same will happen again. More critically, cloud giants like Microsoft, Google, and Amazon are racing to develop proprietary optical interconnects—Google’s Lyra, Microsoft’s Silica—to replace third-party modules. If even 10% of their data center switches to internal solutions by 2028, Zhongji’s revenue projections collapse. This is the same pattern I observed in 2019 with DeFi liquidity: the top 50 high-frequency wallets accounted for 80% of manipulated volume. Here, the top 4 hyperscalers represent over 70% of optical module demand. Customer concentration is a systemic risk, not a strength. The takeaway for crypto investors is clear. The bull market in AI hardware is masking technical fragility, much like the crypto bull market of 2021 masked the weakness of Terra and Three Arrows Capital. The path forward is to position for a cycle that will peak before 2027, not after. Look for signals of deceleration: a single quarter of missed guidance from Nvidia, a regulatory crackdown on AI data center energy consumption, or a breakthrough in alternative networking technologies like CXL or optical compute. When those signals appear, the optical mirage will dissolve. And those who confuse liquidity with settlement will be left holding the bag.