The last time a sell-side analyst published a three-year profit forecast with 100% annual growth, it was 2020, and the stock was Tesla. The market believed the narrative. Then the narrative broke.
Today, Goldman Sachs has done the same for Zhongji Xuchuang, a Chinese manufacturer of optical modules that power AI data centers. The report, leaked through a crypto-native media outlet, predicts 65%, 108%, and 119% net profit growth for 2026, 2027, and 2028 respectively. The implied price target offers 163% upside.
But this is not a story about AI. This is a story about the crypto market's hunger for hardware narratives, the liquidity cycle of institutional money, and the moral hazard of turning a component supplier into a bet on the future of intelligence.
Let me take you back to 2017. I was auditing smart contracts for ICOs, reverse-engineering token distribution mechanisms that promised 'revolution.' One project claimed to be a 'decentralized AWS.' Their whitepaper had a slide comparing themselves to Nvidia. They raised $40 million. The code had a governance backdoor that allowed the team to mint unlimited tokens. The lesson: every infrastructure narrative in crypto is a mirror of the real economy, but with faster hype cycles and fewer regulators.
Now, in 2026, the narrative has shifted from decentralized compute to centralized AI compute. Zhongji Xuchuang is not a crypto company. It does not issue tokens. It manufactures 800G and 1.6T optical modules for hyperscale clouds. Yet the report surfaces on a blockchain news site. Why? Because crypto capital needs a story. The ICO boom died. The DeFi summer faded. The NFT market collapsed. The new alpha is 'AI infrastructure' — but retail cannot directly invest in Nvidia at $2 trillion valuation, so they look for proxies.
Follow the money, not the noise. The money is flowing into AI hardware. The noise is the Goldman Sachs report. The real question is: who benefits from this noise?
Goldman's logic is straightforward: AI training clusters require massive inter-node bandwidth. As GPU shipments grow, so does demand for optical transceivers. Zhongji Xuchuang is the market leader in 800G modules and a primary supplier for Nvidia's 1.6T roadmap. Therefore, revenue and profit will compound at triple-digit rates.
The assumptions are crystalline: - AI capex from hyperscalers grows at >30% CAGR through 2028. - 1.6T module adoption begins in late 2025 and ramps smoothly. - Pricing remains stable, with ASPs rising as technology upgrades. - Zhongji maintains its market share against competitors like Coherent and Eoptolink. - No supply chain disruptions in laser chips, DSPs, or SOI wafers.
The report uses the phrase 'superior execution' three times. It compares Zhongji to Taiwan Semiconductor in the optical domain. It never mentions the word 'risk.'
Here is what the report omits.
First, the optical module industry is notoriously cyclical. In 2019, during the 4G-to-5G transition, demand surged then crashed. Inventory gluts destroyed margins. The same pattern occurred in 2022 after the COVID data center buildout. The current AI-driven demand is real, but it is not immune to cycles. If hyperscalers pause or redirect capex — say, due to efficiency gains from new architectures like CXL — the revenue cliff will be steep.
Second, the customer concentration risk is extreme. Zhongji's top five customers likely account for over 70% of revenue. Nvidia itself is a single point of failure. If Nvidia decides to vertically integrate optical interconnects — it has already acquired Mellanox and is developing its own networking — or if it diversifies to a second supplier, Zhongji's growth trajectory breaks.
Third, the competitive landscape is heating up. Coherent is ramping 1.6T. New entrants in China are gaining traction. Price wars are inevitable. The history of hardware shows that ASP declines often outpace volume gains, squeezing profitability. Goldman's model assumes the opposite.
Fourth, and most important for crypto readers: the report is a classic 'pump the proxy' strategy. The source — a blockchain media outlet — has a history of publishing bullish takes on stocks that align with tokenized narratives. This is not analysis. It is marketing. The target audience is not institutional investors. It is retail traders who have been burned by low-cap altcoins and are desperate for a 'safe' bet.
Volatility is the tax on impatience. The report creates an illusion of certainty. Triple-digit growth for three consecutive years is a statistical anomaly. It happens in hyper-growth phases of a platform company, not in a component manufacturer. Even Nvidia's growth is decelerating as the base effect kicks in. The idea that a module supplier can grow faster than the chip supplier is a red flag.
Now, let me connect this to the crypto infrastructure thesis. Over the past two years, we have seen a wave of projects tokenizing compute resources — Akash, Render, io.net, and others. The narrative is that decentralized compute will compete with centralized clouds. But the reality is that the bottleneck for AI inference is not compute supply; it is interconnect bandwidth. Optical modules are the hidden bottleneck. If you believe in decentralized AI, you must believe that the demand for high-speed optical modules will only increase, regardless of centralization. That part is true.
But the tokenized compute projects do not use 1.6T optical modules yet. They operate on consumer-grade GPUs with standard networking. The Goldman report is about a different market: the top 1% of AI clusters. The market that crypto cannot touch. The market that requires billions of dollars of upfront capex. This is the market that Zhongji serves.
So if you buy the stock based on this report, you are not betting on crypto. You are betting on hyperscalers. You are betting on Nvidia. You are betting on the same centralized forces that crypto claims to disrupt. That is the irony.
What are the signals to watch, then? Not the price target. Not the profit forecast.
First, watch Zhongji's gross margins. If they decline sequentially, the pricing pressure is already here. Second, watch Nvidia's networking revenue. It reports optical module purchases indirectly. If Nvidia's networking ratio drops, it indicates internal development. Third, watch for any announcement from Microsoft or Google about their own optical interconnects. Both have internal projects. Fourth, watch the semiconductor capital equipment orders for VCSEL and silicon photonics. That is a leading indicator of supply.
The time to buy is not when the analyst upgrades. It is when the market panics and the fundamentals remain. In 2022, when data center stocks crashed due to inventory corrections, the right move was to accumulate. Now, with a 163% upside priced in, the risk is asymmetric. The stock already reflects high expectations. Any disappointment will be punished hard.
Let me be direct. This article is a commentary trap disguised as research. It generates emotional conviction in a single narrative. It ignores the counterarguments. It uses authority (Goldman Sachs) to override critical thinking. The blockchain news source adds a veneer of 'future-forward' thinking, but it is the same old trick: create demand for a story, then sell the story.
In 2019, I wrote a piece on the 'Infrastructure Narrative Trap' after watching countless token projects promise to be the 'AWS of blockchain.' Most failed because they built before demand existed. Today, the demand for AI compute is real. But the investment case for Zhongji is not a blockchain case. It is a legacy tech case. And legacy tech cycles are brutal.
The tide does not ask for permission. But it does retreat. When it does, the companies with the strongest balance sheets survive. Zhongji has one. But the stock at current levels is pricing in a tide that never ebbs. It will ebb. The question is when, and by how much.
So here is my takeaway for the crypto-native reader. You are used to risk. You are used to volatility. You have survived 80% drawdowns. The stock market feels safe by comparison. But safety is an illusion. The same forces that drive this narrative — liquidity, greed, and the need for a new story — drive every cycle. The Goldman report is a lighthouse built on sand. The foundation is firm today. The erosion is invisible. But it is happening.
If you want exposure to AI infrastructure, buy the underlying technology: copper, fiber, optical components. Better yet, learn the technology. Understand what makes a 1.6T module different from 800G. Understand the role of DSPs and silicon photonics. Then you can make your own bet, not someone else's.
I hold no position in Zhongji Xuchuang. I have no relationship with Goldman Sachs. I am a researcher who has watched too many narratives inflate and collapse. This one feels different because the underlying demand is real. But the valuation is already discounting paradise. And paradise, in finance, is always a mirage.
Follow the money, not the noise. The money is flowing into AI. But the noise is telling you it will flow forever. That is the trap. Do not step into it.