Why did Qualcomm surge 13% after hours? How does the AI data center strategy drive a valuation revaluation?

On June 24, 2026, Qualcomm held its 2026 Investor Day in New York City. The significance of this event far exceeded that of a routine earnings briefing—it was a systematic declaration by a company renowned for its mobile phone chips for over three decades, positioning itself as a full-stack player in AI infrastructure.

The capital market responded swiftly. After Qualcomm's stock fell 3.29% to $197.41 during regular trading on June 24, it surged over 13% in after-hours trading to $223.56. On June 25, Qualcomm closed at $204.90, up 3.79% from the previous day, with an intraday high of $219.43. Morgan Stanley raised its price target from $146 to $231, a 58% increase, with analyst Joseph Moore admitting that the firm's "skepticism has been wrong all along." Rosenblatt raised its target from $190 to $265 and reiterated a "Buy" rating, calling this Investor Day a "defining turning point" for the company.

Why did the market assign such a positive valuation? The answer points to a core narrative: Qualcomm is transitioning from a valuation framework based on the mobile chip cycle to a growth framework centered on AI inference chips.

Financial Targets Doubled: Non-Phone Revenue Raised to $40 Billion

The most direct signal from this Investor Day came from the substantial upward revision of financial targets. Qualcomm raised its non-phone revenue target for fiscal year 2029 from $22 billion (set 18 months earlier) to $40 billion, nearly doubling it. The compound annual growth rate target for fiscal years 2025 to 2029 is 40%, with a non-GAAP EPS target for fiscal 2029 set at greater than $18.

Among the various business segments, the data center business shows the steepest growth curve. Qualcomm expects data center revenue to reach $5 billion in fiscal 2027, with more than $1 billion each from custom chips for two hyperscale customers. By fiscal 2029, the data center revenue target further jumps to over $15 billion. Going from $5 billion to $15 billion in just two years indicates that Qualcomm's growth curve for the data center business is extremely steep.

Prior to the Investor Day, Bank of America analysts raised their price target for Qualcomm from $165 to $195 but maintained an "Underperform" rating, citing that the company is "entering a fast-growing but fiercely competitive AI market where several large incumbents already exist." This rating itself is a restrained endorsement of Qualcomm's strategy—directionally correct, but execution risks cannot be ignored.

Meta Strategic Partnership Lands: Dragonfly C1000 Gains Key Customer Endorsement

Qualcomm fully disclosed its data center strategy for the first time during the Investor Day, branding it as "Dragonfly." The product portfolio covers four core areas of AI data center infrastructure: connectivity (800G/224G/448G DSP), custom chips (orders won from two hyperscale customers), AI accelerators (AI250 planned for mid-2027), and CPUs (Dragonfly C1000 planned for mid-2028).

The most notable element was the strategic partnership with Meta. Qualcomm announced a multi-year, multi-generation cooperation agreement with Meta, whereby Meta's next-generation servers will adopt the Dragonfly C1000 data center CPU. In a video message during the Investor Day, Meta CEO Mark Zuckerberg stated that under the strategic cooperation agreement, Qualcomm will become Meta's data center CPU supplier. The Dragonfly C1000 features a clock speed exceeding 5 GHz (over 30% faster than competitors) and more than 250 cores, with mass production expected in the second half of 2028.

In addition to Meta, Microsoft will also adopt Qualcomm's High Bandwidth Compute chip architecture for Azure infrastructure. Qualcomm has also secured commitments from two other unnamed hyperscale cloud service providers to use its custom chips. More than 35 global supply chain companies, including Compal, Delta Electronics, Foxconn, Quanta Computer, UMC, and Nanya Technology, have expressed support for Qualcomm's data center vision.

Acquisition of Modular for ~$4 Billion: Fills the Gap in AI Inference Software Ecosystem

Beyond hardware, Qualcomm is simultaneously building its software ecosystem. Qualcomm announced the acquisition of AI software startup Modular in an all-stock transaction valued at approximately $3.92 billion, expected to close in the second half of 2026.

Modular's core value lies in its ability to run AI models across different chips without requiring developers to write separate code for each processor. This acquisition is seen as a direct response to Nvidia's CUDA ecosystem. CUDA has formed a formidable moat by locking in millions of developers, and Modular's software technology is expected to help Qualcomm break through this barrier.

In his opening remarks, Qualcomm President and CEO Cristiano Amon defined the company's next chapter: "We are accelerating our edge diversification strategy, unveiling a comprehensive roadmap for next-generation AI data centers, and evolving into a platform company." The full-stack layout from hardware to software is a concrete manifestation of this platform strategy.

Logic of Valuation Restructuring: From Mobile Cycle to AI Inference Growth Curve

The valuation significance of Qualcomm's strategic shift must be understood within a broader industrial context.

Traditionally, the market valued Qualcomm based on the smartphone chip cycle. The smartphone market has entered a phase of stock competition, with major customers like Apple and Samsung increasingly developing their own chips. The slowing growth and cyclical fluctuations of the mobile phone business have limited the expansion of Qualcomm's valuation multiples.

The market space for AI inference chips is entirely different. Bank of America analysts point out that AI inference—running already-trained AI models—has become a key battlefield in the semiconductor industry. Wells Fargo estimates the total addressable market for AI inference chips exceeds $100 billion. Qualcomm expects data center revenue to exceed $15 billion in fiscal 2029, a relatively small share of this vast market—but precisely for this reason, the growth potential is substantial.

From a valuation methodology perspective, the market is applying a triple discount to Qualcomm: first, the stable cash flow from its existing mobile phone business; second, the exponential growth of the data center business from $30 million (fiscal 2026) to $15 billion (fiscal 2029); and third, the option value of long-term market share gains in AI inference chips.

Morgan Stanley raised its price target 58% to $231, Rosenblatt set a target of $265, Bernstein raised its target from $140 to $235, and Citigroup raised its target from $160 to $198. The range of these targets ($198–$265) itself reflects varying levels of confidence in Qualcomm's AI strategy—optimists see a decisive turning point, while prudent investors await execution validation.

Risks and Challenges: Execution Test in a Crowded Arena

Qualcomm's AI data center strategy is not without risks. Although Bank of America raised its price target to $220, it maintained an "Underperform" rating, believing that the current stock price already factors in a significant degree of data center success expectations.

The competitive landscape is the biggest uncertainty. Nvidia holds a near-monopoly in the AI training chip market; Broadcom and Marvell are expanding in custom ASICs; and cloud vendors' in-house chips like Amazon's Graviton and Google's Axion are eroding market space. Qualcomm expects data center revenue of $5 billion in fiscal 2027, while Bank of America analysts estimate data center revenue of approximately $2 billion to $5 billion for fiscal 2027–2028—the upper end of this range is roughly in line with company guidance, but the lower end is only 40% of guidance.

Additionally, Bernstein notes that weakness in the smartphone business could drag near-term earnings before the data center business reaches scale. Mobile phone revenue is expected to shrink to about one-third of total revenue by fiscal 2029, but until then, Qualcomm needs to maintain its mobile business base during the transition.

Crypto Market Perspective: Tech Stock AI Narrative and Capital Competition with Risk Assets

As of June 26, 2026, Bitcoin is trading around $59,400–$59,700, down about 2.86% in 24 hours, officially breaking below the $60,000 mark. Ethereum is around $1,560, dropping nearly 5% in 24 hours. The total crypto market cap has shrunk to approximately $2.06 trillion from early 2026 highs. The Fear and Greed Index has dropped to 18, deep in extreme fear territory.

This crypto market decline is structurally linked to the capital suction effect of AI chip stocks. According to the head of research at CF Benchmarks, recent incremental capital and investor attention have flooded into AI concept stocks, leaving cryptocurrencies to compete for a smaller share of overall risk appetite. Bitcoin's correlation with the Nasdaq is as high as 0.94, with capital flows in the U.S. tech sector directly affecting the crypto market.

Qualcomm's strategic leap from mobile phone chips to AI inference chips perfectly sits at the intersection of this industrial transformation. AI infrastructure construction is becoming a core direction of global capital allocation, and whether Qualcomm can complete its valuation restructuring from a "mobile chip company" to an "AI infrastructure company" will depend on the actual shipments and customer deployment progress of its Dragonfly products over the next 12–24 months.

Conclusion

Qualcomm's Investor Day on June 24, 2026, marks the official launch of the most significant strategic transformation in the history of this over-three-decade-old semiconductor company. From mobile chips to AI inference chips, from consumer electronics to data center infrastructure, from hardware supplier to full-stack platform company—Qualcomm is redefining its position in the industry.

The market has given an initial positive valuation, but true value restructuring requires product delivery and revenue realization for validation. Meta's orders, Microsoft's collaboration, and Modular's acquisition provide verifiable logical anchors for Qualcomm. In the AI inference chip TAM of over $100 billion, whether Qualcomm can capture a meaningful share will determine whether its valuation migrates from mobile chip's 20x P/E to a higher multiple for AI infrastructure, or reverts to cyclical stock pricing logic.

For crypto market investors, Qualcomm's story also serves as a reference—as AI infrastructure construction becomes the core narrative of global capital allocation, the capital flows and valuation logic of risk assets are being rewritten.

FAQ

Q1: What are the specifics of the strategic partnership between Qualcomm and Meta?

Qualcomm announced at the 2026 Investor Day a multi-year, multi-generation cooperation agreement with Meta. Meta's next-generation servers will adopt Qualcomm's Dragonfly C1000 data center CPU. The Dragonfly C1000 is planned for mass production in the second half of 2028, with a clock speed exceeding 5 GHz and more than 250 cores. Additionally, Microsoft will adopt Qualcomm's HBC chip architecture for Azure infrastructure.

Q2: What financial targets did Qualcomm raise?

Qualcomm raised its non-phone revenue target for fiscal 2029 from $22 billion to $40 billion. Among them, the data center business target exceeds $15 billion, the automotive business is $10 billion, and the IoT business exceeds $14 billion. The compound annual growth rate target for fiscal 2025 to 2029 is 40%, and the non-GAAP EPS target for fiscal 2029 is greater than $18.

Q3: What is the significance of Qualcomm's acquisition of Modular?

Qualcomm acquired AI software startup Modular in an all-stock transaction valued at approximately $3.92 billion. Modular's software enables AI models to run on different chips without writing separate code for each processor. This acquisition aims to challenge Nvidia's CUDA ecosystem's developer lock-in advantage and strengthen Qualcomm's capabilities in the AI inference software layer.

Q4: What are the latest ratings from Wall Street institutions on Qualcomm?

Morgan Stanley upgraded Qualcomm from "Underweight" to "Equal-weight" and raised its price target by 58% from $146 to $231. Rosenblatt raised its target from $190 to $265 and reiterated "Buy." Bernstein raised its target from $140 to $235. Bank of America raised its target from $165 to $220 but maintained "Underperform."

Q5: What are the main risks facing Qualcomm's AI data center business?

Main risks include: Nvidia's near-monopoly in AI training chips, competition from Broadcom and Marvell in custom ASICs, and substitution effects from cloud vendors' in-house chips like Amazon's Graviton and Google's Axion. Additionally, Bernstein notes that smartphone business weakness could drag near-term earnings before the data center business reaches scale. Bank of America believes the current stock price already factors in a significant degree of data center success expectations.