Lenovo CFO: Company 'Completely Undervalued' by Market, Revenue Could Hit $100B Ahead of Schedule

The Detail That Stuck

After attending an AI investment industry summit this week, the most memorable takeaway wasn't any speaker's forecast—it was a detail from a factory floor.

Chen Zhenkuan, Vice President of Lenovo Group, shared during a panel discussion that their factory producing supernodes is undergoing upgrades. Traditional server production lines could be placed on the second or third floor, requiring floor load capacity of just one ton per square meter. For supernodes, the minimum requirement is three tons, designed with a margin of five tons. This means production lines must move to the ground floor, and experts insist on solid earth foundations—factories with underground parking garages simply cannot handle the job. Going further, each supernode production base requires over 30 megawatts of power, necessitating a dedicated 110 kV substation.

While capital markets debate model ceilings and valuation percentiles, the industry is recalculating concrete load-bearing capacity. For us secondary market investors, the key insight is this: AI infrastructure is shifting from "buying equipment" to "building systems." Systemic changes ultimately manifest in things this concrete, this heavy.

01 The New Narrative Standard for AI

Professor Peng Wensheng from Shanghai Advanced Institute of Finance offered a framework at the summit that I found more useful than most AI narratives: Large language models represent "scalable production of cognition," but unlike the asset-light internet platforms of the past, every step forward for models requires real capital investment in computing power, data, and electricity. AI is pushing the digital economy into an industrialization phase with significantly front-loaded capital expenditure.

The numbers make it clearer. In Q1 2026, the combined quarterly capital expenditure of Amazon, Microsoft, Google, and Meta—the four North American cloud giants—reached $131.6 billion, up 70% year-over-year, with a full-year estimate of approximately $710 billion. Wang Qing of Chongyang Investment pushed further: U.S. leading firms' AI capital expenditure was about $400 billion last year, could double to $800 billion this year, and may reach $1 trillion next year. China is slower, with about RMB 500 billion last year and RMB 1 trillion this year.

At this scale, the rules change. Wang Qing highlighted an easily overlooked variable: starting this year, these giants' free cash flow has turned negative, prompting large-scale bond issuance and off-balance-sheet financing. Projects with major corporate credit backing carry financing rates of 5.3% to 6.6%; those without, 8.5% to 9%. For the first time, the technology cycle and financial cycle are so tightly intertwined. Liu Yuhui's assessment was even blunter: AI is approaching the "intelligence singularity," and this is precisely the most capital-intensive phase.

Thus, the core question for this AI trading cycle has shifted from "Does the story make sense?" to "Can the money be recovered?" In other words: orders prove demand, profit margins prove business models, and cash flow determines how much industry prosperity translates into company value. After more than three years of rapid AI investment growth, the narrative logic for the next phase will likely follow this standard.

02 Lenovo as an Individual Reference Point

By this standard, Lenovo is a ready-made case study.

According to the latest IDC data: In Q2 2026, Lenovo shipped 286,000 x86 servers globally, up 45.6% year-over-year, ranking first worldwide. By revenue, it grew 96% year-over-year to $8.26 billion, just 0.6 percentage points behind Dell. In the latest fiscal quarter, the Infrastructure Solutions Group (ISG) revenue grew 98% year-over-year to RMB 57.9 billion, with an operating profit margin of 9.1%—an all-time high. AI server backlog orders more than doubled in a single quarter, reaching over RMB 360 billion. Management added a crucial note: most of these orders are deliverable within one year, contingent on supply chain readiness.

There's a telling detail in the order structure. One new cloud client started with a budget of just a few hundred million dollars, which grew to tens of billions within nine to ten months. Such ramp-up speed for a single client was unprecedented in the traditional IT era.

CFO Ken Zheng gave an honest accounting during a summit dialogue, not shying away from the thin margins of AI servers: New cloud clients have gross margins of 25% to 30%, with some overseas scenarios higher; enterprise clients have gross margins above 30% and operating profit margins up to 18%; overseas business deeply tied to NVIDIA faces margin pressure, relying on scale. The path to net profit margins of 5% or even 8%+ doesn't lie in any single blockbuster product but in revenue structure rebalancing: storage business has 60% to 70% gross margins, with a long-term target of $5-6 billion in revenue and $1 billion in profit; the services business includes equipment leasing and high-margin pure software increments. The "hardware plus services" synergy, according to management, is something the market hasn't fully priced yet. Further out is enterprise-level inference, which Zheng calls "the opportunity for AI over the next decade, still in its early stages."

The timeline is also laid out: one to two years to reach $100 billion in revenue (which he believes may be achieved ahead of schedule) with net profit margins above 3%; three to five years to $130 billion with net margins above 5%; five years and beyond to $150 billion with net margins above 8%.

03 How the Market Cap Equation Works

Beyond the scorecard, I want to deconstruct Lenovo's market cap logic from the mechanism perspective of a "global AI infrastructure leader." Market cap is essentially three numbers multiplied: Revenue × Net Profit Margin × Valuation Multiple. Each variable has its own driving mechanism.

Revenue is driven by market share and delivery scheduling. Being the global leader in shipments isn't just a title—it means scale advantages in procurement, manufacturing, and delivery, which in turn support further market share gains. Wang Qing's estimate: the global server market in 2026 is about $700 billion, with the U.S. at $500 billion and China at $100 billion. Lenovo is already second globally by revenue, just 0.6 percentage points behind Dell. Every percentage point of market share movement represents tens of billions in revenue. And the RMB 360 billion backlog, mostly convertible within a year, turns "growth expectations" into "growth schedules"—which carries more weight for valuation than any long-term story. ISG's single-quarter revenue of RMB 57.9 billion annualizes to about $32 billion, not yet including the long-term $5-6 billion from storage and services growth.

Net profit margin is driven by structural quality. ISG's OPM has reached 9.1%, but it's a mix of three business types: thin-margin new cloud deals, high-margin enterprise clients (OPM 18%), and even more profitable storage (60% to 70% gross margins) and services. As the structure shifts toward the latter two, net margins will climb from the current 2-3% range toward 3%, 5%, and 8%. Two amplifiers support this: first, being the top shipper creates cost advantages in procurement and manufacturing; second, supernodes are locking out "pure server vendors," concentrating competition among full-stack capable leaders, improving pricing power. Structure, scale, and barriers—three forces moving in the same direction.

Valuation multiple is driven by a shift in the valuation framework. The market assigns one multiple to a hardware manufacturer with 2% net margins and another to an "AI systems plus services" company—Dell enjoys the latter. Lenovo's current valuation is roughly one-seventh of Dell's, "but our revenue and net profit aren't seven times different," as Zheng put it. This gap breaks down into three discounts: the historical label as a PC company, the low-margin manufacturing identity, and market concerns about order quality (new cloud client creditworthiness). These three factors are being disproven quarter by quarter through order conversion and margin improvement.

Multiplying these three factors yields a rough calculation (note: not a forecast): $100 billion in revenue at 5% net margins equals $5 billion in net profit. At that point, if the market still applies a "low-margin hardware" multiple, market cap growth mainly comes from profit expansion. If the multiple shifts toward systems and services providers, both profit and multiple expand in tandem. This is why the market cap curve for such assets is often not a steady upward slope but step-like—each time net margins cross a threshold (3%, 5%), the market re-rates the valuation framework. Even if the current 1/7 gap narrows by half, the upside is more than three times. Zheng's confidence that "we are completely undervalued by the market" likely stems from this: all three variables have clear mechanisms driving them, rather than betting on a trend.

A caveat: multiple migration depends on timely order conversion and margin improvement. If any link breaks, all three factors contract together. This is why the market now only pays for what can be recovered.

04 From Stacking Chips to System-Level AI Sovereignty

Returning to industry logic, supernodes represent the hardest piece of this puzzle and carry the most significance for China's industrial policy.

The reasoning is straightforward. DeepSeek's 671 billion parameters can run on a single 8-card cabinet. Kimi 3.5, with 2.8 trillion parameters, already strains an 8-card cabinet. The next generation of 10 trillion-parameter models will require 32-card supernodes or even clusters to operate.

Chen Zhenkuan's view: Supernodes are "not a choice but a necessary AI infrastructure." His timeline: about five AI chip companies will support supernodes this year, with all leading players capable by next year—"Next year will definitely be the big year for supernodes. This year is still about feasibility; next year will see large-scale commercial deployment." Lenovo's own product line is accelerating: a 40-card inference product has been launched, with 128-card and 256-card models under development, with prototypes expected by end of 2026 capable of supporting 10,000-card training clusters. The company has already won a bid from a major domestic internet firm, with delivery in the second half of the year.

Supernodes fundamentally change the nature of competition. A single supernode cabinet must be divided into computing cabinets, power cabinets, storage cabinets, and connection cabinets. Optics, liquid cooling, AI storage, vertical and horizontal scaling—all must work within the same system. Chen put it bluntly: "A pure server vendor today cannot make a supernode. You need to integrate teams for servers, storage, networking, operating systems, and software to define a single product. Comprehensive capability, all-around capability—this is a certainty."

This statement must be viewed within the context of U.S.-China tech competition. The gap in advanced process nodes is unavoidable in the short term—that's a fact. But computing competition has never been a single path. When single-point process technology can't catch up, densifying inter-card connectivity, expanding cache, and improving system-level efficiency—using engineering and integration to offset single-chip gaps—is a path China must and can take under current constraints. The country's software engineer talent pool, world-leading manufacturing and delivery capabilities, and relatively abundant power supply all align with this approach.

Chen offered a comparison at the summit: The U.S. faces power shortages; China faces chip shortages. The power supply capacity of a single Ulanqab data center hub equals that of six average U.S. states. With a $700 billion global server market—$500 billion in the U.S. and $100 billion in China—China's share shouldn't be just one-fifth of America's. This gap itself represents growth space for the coming years.

The policy implications are already emerging: AI infrastructure competition is no longer about chips alone but about comprehensive competition in power, grid, land, industrial standards, and systems engineering. When supernode production lines require five tons per square meter of floor load, 30 megawatts per base, and dedicated 110 kV substations, "computing sovereignty" moves from concept to planning documents—power planning, industrial land use, and manufacturing upgrades all become fronts for AI infrastructure policy.

Chen also flagged a variable: different chip manufacturers each define their own interconnect protocols. "Twelve manufacturers have 13 protocols." Without standardization, system costs across the industry will rise. This is both a business issue and a question of standard-setting influence, requiring industrial coordination to resolve.

05 The Decade of Inference

Another consensus from the summit concerned timing.

Song Yuchun of Lenovo Capital's assessment: AI represents the early stage of the Fourth Industrial Revolution. Generative AI has truly exploded for less than four years, and the market has mainly seen the first wave of infrastructure buildout. "Actual penetration into thousands of industries hasn't really begun." His Lenovo Capital has invested in over 330 companies and 26 IPOs over a decade, with more than 150 in AI—spanning chips, infrastructure, large models, agents, autonomous driving, and embodied intelligence. He says AI has been the best-performing sector in terms of deal count, investment amount, and returns over this decade, and the best may still be ahead.

On the enterprise side, Zheng's observation was more specific: Large international companies are still in the early stages of AI adoption. They must first address internal data, organizational structure, and business process integration. Companies that have undergone mergers and acquisitions face system integration that can take years. But once these foundational tasks are completed, enterprise-side inference demand will be substantial. Combined with data security driving clients toward local private deployment, enterprise computing budgets are expected to be released in a concentrated manner around 2027.

He also shared a telling detail. When overseas CFOs discuss AI optimization, their first reaction is often "we need to cut 25% of staff." His perspective is different: if revenue reaches $150 billion with net margins of 5 to 8 points, "I wouldn't just avoid layoffs—I'd hire more people. But I'd hire AI talent." In the public debate about AI and employment, this is one of the few concrete, quantified counter-arguments from an operator's perspective.

06 Conclusion

Risks are certainly on the table: financing costs, crowded trades, and periodic upstream component price increases. Summit investors didn't avoid these topics. Wang Qing's reminder was sobering: the evolution of fundamentals and capital market pricing are not always synchronized.

But return to that five-ton floor. Every industrial revolution in history has not manifested through keynote speeches but through foundations, power grids, steel beams, and load-bearing calculations.

Today's AI has reached that point.

Disclaimer: The above content (including images and videos if any) is uploaded and published by a user of NetEase Hao, a social media platform that provides information storage services only.