Qualcomm Targets $15 Billion Data Centre Push as Mobile Chip Sales Slide

A Strategic Shift Under Pressure

Qualcomm has outlined an aggressive expansion into AI data centre infrastructure, projecting that the segment will generate $15 billion in annual revenue by 2029 and account for roughly 30 per cent of total company sales. The move comes as the chipmaker confronts a sharp contraction in its traditional smartphone business, where sales have fallen approximately 20 per cent compared to 2025 levels.

During its fourth-quarter 2026 earnings discussion, Qualcomm disclosed that it expects data centre chip revenue to reach $650 million in the first quarter of 2027, climbing to $5 billion by the end of that year. The trajectory represents one of the fastest pivots in the company's recent history, driven by partnerships with hyperscale operators including Amazon and Meta that were formalised earlier in 2026.

Against total company revenue of $10.4 billion in 2026, the projected data centre figures would mark a fundamental reshaping of Qualcomm's business mix. The firm has historically derived the bulk of its income from mobile processors and licensing fees tied to cellular standards, a revenue stream now under pressure from both market saturation and lengthening device replacement cycles.

The Economics of the Pivot

Qualcomm's 30 per cent target sits well below the ambitions of rival AMD, which has indicated that data centres will eventually contribute more than 80 per cent of its revenue. Yet the divergence reflects differing starting points rather than differing assessments of opportunity. AMD entered the data centre market earlier and has already built substantial server CPU and GPU businesses; Qualcomm is still establishing its foothold in inference and edge AI workloads where its low-power design expertise offers differentiation.

The company's partnerships with Amazon and Meta provide anchor customers for its data centre product line, offering both validation and scale. Amazon's infrastructure requirements span general-purpose compute and machine learning inference, while Meta's focus on large-scale AI training and recommendation systems creates demand for chips optimised for specific workload profiles. Qualcomm has positioned its data centre offerings around power efficiency and cost per inference, areas where its mobile heritage in thermal management and energy optimisation translates directly.

The smartphone decline that prompted this shift is rooted in macroeconomic headwinds and market maturity. Replacement cycles in key markets including China, India, and Southeast Asia have extended beyond three years, while consumer spending on discretionary electronics has softened. Qualcomm's reliance on Android device makers, who face intensifying competition from local chipmakers in China and India, compounds the pressure.

Infrastructure Demand and Competitive Landscape

The data centre pivot rests on sustained demand for AI compute capacity, particularly inference workloads that run trained models in production environments. Inference represents a growing share of total AI spending as models move from research labs into consumer applications, enterprise software, and edge devices. Qualcomm's approach targets this segment rather than the training market dominated by Nvidia's high-performance GPUs.

Competitive intensity in data centre silicon is rising. Nvidia holds commanding market share in AI training accelerators, while AMD, Intel, and a cohort of startups including Cerebras and Groq are vying for inference and alternative architecture niches. Hyperscale cloud providers are also designing proprietary chips, with Amazon's Graviton and Trainium, Google's TPU, and Microsoft's Maia reducing reliance on merchant silicon vendors.

Qualcomm's entry hinges on its ability to offer differentiated economics, particularly in power consumption and total cost of ownership. Data centre operators face mounting electricity costs as AI workloads scale, making energy efficiency a critical purchasing criterion. The company's mobile chip design methodology, which prioritises performance per watt, positions it to compete on this dimension even as it builds out software ecosystems and developer tooling required for enterprise adoption.

Financial Stakes and Execution Risk

The revenue ramp Qualcomm has outlined is steep. Growing from $650 million to $5 billion within a single year requires not only customer commitments but also supply chain scaling, manufacturing capacity, and successful silicon tape-outs. The company relies on third-party foundries, primarily Taiwan Semiconductor Manufacturing Company, for production, exposing it to capacity allocation decisions and geopolitical risk around Taiwan.

Execution risk extends beyond manufacturing. Data centre customers demand multi-year roadmaps, software maturity, and ecosystem support that Qualcomm has not historically provided at the same scale as established server vendors. The company must build relationships with cloud software providers, AI framework developers, and systems integrators while competing for engineering talent in a market where AI specialists command significant premiums.

The 30 per cent revenue target implies that Qualcomm's overall business will grow substantially, or that mobile and other segments will continue to contract. If total revenue remains near current levels, a $15 billion data centre segment would require company-wide revenue of $50 billion, more than quadruple 2026 figures. Alternatively, if mobile revenue stabilises and adjacent segments like automotive and Internet of Things grow modestly, the data centre business could reach the 30 per cent share at lower absolute scale.

What It Signals for the Sector

Qualcomm's commitment underscores how thoroughly AI infrastructure spending is reshaping semiconductor industry strategy. Companies once defined by consumer electronics or PC markets are redirecting capital and engineering resources toward data centre products, betting that the current wave of AI investment will persist and expand.

The firm's first-quarter 2027 revenue guidance, expected in early January, will offer an initial test of whether customer deployments are tracking to plan. Qualcomm has scheduled a data centre product briefing for February 2027, where it is anticipated to detail chip architectures, performance benchmarks, and customer case studies.