This piece breaks down Qualcomm, a company that makes money more from patent licensing than selling chips. The author says Qualcomm's real moat is that any phone connecting to a network must pay it royalties. But risks include Apple (its biggest customer, now trying to build its own modem, likely not ready until 2025), MediaTek (a rival catching up), and Meta (bullish, dropped its own VR chip for Qualcomm's). Also noted: 60-70% of sales go to China, a geopolitical risk.
Qualcomm is a semiconductor company with an enterprise value of $150 billion, dominating the 3G, 4G, and 5G mobile communication standards. The report's core argument holds that Qualcomm builds a strong moat through standard-setting and chip design (rather than manufacturing), and its business model
Jay Goldberg (semiconductor industry advisor, partner at Snowcloud Capital) breaks down Qualcomm — a semiconductor company with an enterprise value of $150 billion that dominates 3G/4G/5G mobile communication standards. Goldberg argues that Qualcomm's true moat lies not in chip manufacturing, but in its patent licensing model established through standard-setting — "if you want to make a phone that can communicate with telecom operators, you must obtain a license from Qualcomm" — which generates consistently high-margin revenue.
Jay Goldberg argues that Qualcomm’s history is essentially a story of victory in the "wireless standards wars," rather than a simple growth story of a semiconductor company.
Qualcomm was founded in 1985 by Irwin Jacobs and Andrew Viterbi, who had previously co-founded Linkabit in 1968 (later sold to Maycom), one of the companies with the most spin-offs in Silicon Valley. Qualcomm’s early product, Omnitracks, provided satellite tracking services for long-haul trucks, but the real goal was to invest profits into the development of mobile communication standards.
The key turning point came in the 3G era. In the late 1980s, the European-dominated GSM standard held 70% of global subscribers, while Qualcomm’s CDMA standard, though technically superior (Goldberg cautiously notes that "saying this in certain bars could start a fight"), never captured more than 20% market share. By deeply engaging with the United Nations-affiliated wireless standards-setting organizations, Qualcomm incorporated its CDMA technology extensively into global 3G standards. "From that point on, virtually every mobile phone in the world has used Qualcomm’s technology or IP in some way — and this situation continues to this day."
Goldberg points out that Qualcomm has two distinct yet mutually reinforcing revenue streams: the semiconductor business (QCT) and the licensing business (QTL), with the latter contributing nearly 30% of operating profit from just 15% of revenue.
| Business Segment | Revenue Share | Gross Margin | Core Mechanism |
|---|---|---|---|
| QCT (Chips) | ~85% | ~60% | Sells baseband/modem chips to enable communication between phones and carriers |
| QTL (Licensing) | ~15% | 85-90% | Collects patent royalties of 2%-5% of handset selling price |
The logic of the licensing business: Qualcomm is legally required to offer patent licenses to all manufacturers, but the law does not specify royalty rates. Qualcomm charges 2%-5% of the selling price per phone. "This is the monetization of their 30 years of R&D investment," Goldberg emphasizes. This model allows Qualcomm to generate a sustained stream of high-margin cash flow beyond chip sales.
The cyclical nature of the chip business: Qualcomm typically enjoys a 1-2 year "quasi-monopoly" at the onset of each new standard generation (3G→4G→5G)—it is the sole supplier of next-generation baseband chips in the market. However, as competitors (such as MediaTek) catch up, market share is gradually eroded. Between 2015 and 2020, Qualcomm's revenue growth was sluggish, which Goldberg attributes to "smartphone market saturation" and "the end of the standards war."
Goldberg describes the relationship between Qualcomm and Apple as "an irresistible force meeting an immovable object" — both are accustomed to doing things their own way.
Historical Context:
Current Situation and Risks:
Goldberg notes that CEO Cristiano Amon, who took office in 2021, is driving Qualcomm's strategic transformation — no longer calling itself a "wireless company" but an "edge computing company."
The three key directions of the new strategy:
1. Automotive Electronics: Connecting vehicles, digital dashboards, and electric vehicles — "This is a very fitting application scenario for Qualcomm's technology." Qualcomm acquired Arriver, the software business of European automotive supplier Vionier, and integrated it into its automotive solutions.
2. Internet of Things: "This is a very vague concept, not a single market, but a collection of many different markets. It is happening, but it will take time to penetrate."
3. AR/VR: Meta (Facebook) has abandoned its self-developed VR chips and fully shifted to Qualcomm. Goldberg believes that "except for Apple, which may develop its own chips, Qualcomm is in the right position everywhere."
Capital Allocation: Qualcomm generates $6-10 billion in free cash flow annually (operating cash flow of $8-10 billion, with capital expenditures of approximately $2 billion). Goldberg expects the company to primarily return capital through share buybacks and dividends, while also pursuing "billion-dollar-level" bolt-on acquisitions (such as Nuvia for core computing capabilities).
Goldberg believes that Qualcomm’s primary competitive threat does not come from traditional rivals, but from the trend of customers developing their own chips.
Current Competitive Landscape:
| Price Range | Major Suppliers |
|---|---|
| $600+ | Qualcomm dominates |
| $400-600 | Qualcomm competes with MediaTek |
| Below $300 | MediaTek or UNISOC |
Key Risks:
1. Apple’s lead in custom application processors: "Apple’s A-series chips are about two years ahead of Qualcomm’s Snapdragon processors." This is not just a loss of Apple’s business—more critically, Apple’s competitors (Xiaomi, Samsung, OPPO, vivo) need to compete against iPhones powered by Apple’s custom chips, and it remains uncertain whether Qualcomm can deliver solutions with equivalent performance.
2. Geopolitical risks: Qualcomm generates 60%-70% of its sales in China (via contract manufacturers such as Foxconn). "If China chooses to retaliate, Qualcomm would be a very natural target." The company has already faced antitrust and tax issues in China.
3. Standard cycle gap: 5G has been largely deployed, and 6G is not expected until the late 2020s. While there are intermediate technologies like millimeter wave and 5.5G, "it will take a long time to materialize." This means Qualcomm will experience a "prolonged cyclical pause."
4. Foundry dependency: Global advanced process nodes are almost entirely dependent on TSMC. Qualcomm has attempted to maintain bargaining power by using both TSMC and Samsung for foundry services, but "Samsung lags behind TSMC by 6-12 months."
| Position | Analyst View | Key Data |
|---|---|---|
| Apple | Risk warning (largest customer but plans to develop in-house baseband) | Avoided using Qualcomm baseband in the first 12 iPhone generations; settled lawsuit in 2019 due to 5G baseband demand |
| MediaTek | Neutral (competitor, narrowing gap) | Lagged behind Qualcomm by 3 years in the 4G era, only 6 months behind in the 5G era |
| Intel | Risk warning (baseband business failure) | Failed to deliver 5G baseband on time, leading to Apple's settlement |
| Broadcom | Historical event (once launched a hostile takeover) | Launched a hostile takeover in 2017, blocked by the Trump administration on national security grounds |
| NXP | Historical event (failed acquisition) | Qualcomm planned to acquire NXP as a defense strategy against Broadcom, failed due to delays in Chinese regulatory approval |
| Meta | Bullish (abandoned in-house development, turned to Qualcomm) | Has abandoned in-house VR chip development, fully adopted Qualcomm's solution |
| TSMC | Neutral (key supplier) | The only foundry capable of producing advanced process nodes |
| Samsung | Neutral (foundry alternative + customer) | Foundry capability lags behind TSMC by 6–12 months |
1. "Patent everything, and be prepared to defend them in court repeatedly." (Goldberg) — Qualcomm’s headquarters lobby is lined with tens of thousands of patents, intended to create a deterrent. This is the foundation of its licensing business model.
2. "The relationship between Apple and Qualcomm is an irresistible force meeting an immovable object." (Goldberg) — Both companies are accustomed to doing things their own way, leading to a multi-year legal battle that ultimately ended in a settlement due to technological dependence.
3. "Developing an in-house baseband is actually very difficult—it requires a completely different skill set." (Goldberg) — Apple’s in-house baseband plan has been delayed to 2025, and its success remains uncertain. This provides Qualcomm with a window of opportunity.
4. "Qualcomm enjoys a quasi-monopoly for 1–2 years at the launch of each new standard." (Goldberg) — This is the core mechanism behind Qualcomm’s cyclical revenue fluctuations: first-mover advantage generates excess profits, followed by catch-up from competitors.
5. "The biggest threat is not traditional competitors, but the trend of customers developing their own chips." (Goldberg) — Apple’s application processor leads Qualcomm by about two years, and its competitors need alternatives, posing a medium- to long-term risk.
6. "60%–70% of Qualcomm’s sales go to China—if China chooses to retaliate, Qualcomm would be a very natural target." (Goldberg) — Geopolitical risk is the most uncontrollable external factor for Qualcomm.
7. "Cristiano Amon comes from the carrier relations department, reflecting the historical importance of that business to Qualcomm." (Goldberg) — Qualcomm’s success has largely depended on collaboration with telecom operators to drive standard evolution, rather than pure chip sales.
8. "Qualcomm is still doing some truly fundamental engineering innovation—if successful, it could reshape the entire game again." (Goldberg) — Its RF front-end (RFFE) business has the potential to become a second strategic high ground after baseband.