Qualcomm is asking investors to look beyond smartphones. After reporting a 26% drop in fourth-quarter handset-chip revenue, the company said on November 4, 2026, that it still expects its data center business to generate $5 billion in fiscal 2027. The Qualcomm data center revenue target is ambitious for a chipmaker whose latest results remain dominated by phones, but management is counting on custom AI silicon and expanding cloud relationships to change that balance.
The target was reiterated in Qualcomm’s official fiscal 2026 earnings release, not announced as revenue already earned. The company expects non-handset revenue to grow by more than 60% year over year in fiscal 2027, which began after its financial year ended September 27, 2026. Its near-term challenge is to build that new business while absorbing a decline in its established handset market.
Phones drag down the quarter despite stronger automotive sales
Qualcomm recorded $10.322 billion in fourth-quarter revenue under generally accepted accounting principles (GAAP), down 8% from $11.270 billion a year earlier. On a non-GAAP basis, quarterly revenue was $10.442 billion, a decline of 7%. Full-year GAAP revenue totaled $43.120 billion, down 3% from fiscal 2025. Those figures put the company’s diversification plans in context: rapid growth in emerging product lines has yet to reverse the fall in its largest business.
The semiconductor segment, known as QCT, generated $5.137 billion from handsets during the quarter, compared with $6.961 billion a year earlier. Automotive revenue grew 89% to $1.990 billion, while Internet of Things (IoT) revenue increased 4% to $1.888 billion. Overall QCT sales declined 8% to $9.015 billion. The separate technology-licensing segment, QTL, reported $1.304 billion, down 7% year over year.
The contrast is striking. Automotive supplied almost $937 million more revenue than in the comparable quarter last year, but handset sales fell by about $1.824 billion. That arithmetic helps explain why Qualcomm can deliver impressive growth in cars and connected devices yet still report lower total sales. It also shows why a substantial data center business could matter: management needs a new source of scale, not simply another high-growth niche.
Fiscal-year earnings also need careful interpretation. GAAP net income reached $13.628 billion, but the company reported a $5.7 billion income-tax benefit associated with releasing a valuation allowance. That one-off benefit means the sharp increase in GAAP profit should not be mistaken for equivalent growth in underlying chip demand. Fiscal 2026 non-GAAP net income, by comparison, declined from the prior year.
What Qualcomm plans to sell to AI data centers
Rather than relying only on the smartphone processors that made its Snapdragon brand familiar, Qualcomm is pursuing chips and connectivity technology designed for cloud AI workloads. A central opportunity is inference: running a trained AI model to answer questions, generate content or complete tasks. At scale, inference infrastructure must balance compute performance with memory, networking capacity, energy use and cost per request.
In September, Qualcomm and Amazon disclosed a multi-generational collaboration on custom AI data center silicon. Their work covers customized chips for AI inference and optical networking that the companies say could support links at 1.6 terabits per second and future generations. Qualcomm also said it would expand its use of AWS services, including Amazon Bedrock, in electronic design automation workflows. The partnership provides a clearer view of the intended market, although the announcement does not establish how much revenue each product will deliver in a particular quarter.
Qualcomm’s move fits a broader trend among cloud operators: they are seeking chip designs tailored to their own workloads rather than relying on one general-purpose accelerator for everything. Custom silicon can potentially improve efficiency for predictable tasks, but customers still have to evaluate software compatibility, total deployment cost, reliability and networking performance. A promising design or partnership is not the same as an installed fleet of revenue-producing systems.
The company described a larger roadmap at its June 2026 investor event, including a data center revenue goal above $15 billion by fiscal 2029 and a broader ambition to make handsets about one-third of QCT revenue by then. These are management targets extending several years into the future. Fiscal 2027 is a nearer test of whether that diversification strategy can produce measurable sales.
Why the $5 billion target matters for the AI chip market
For buyers of AI infrastructure, another credible supplier of specialized silicon could widen architectural choices. The opportunity is not confined to the processor itself: optical connections, memory subsystems and software determine how efficiently large installations operate. Qualcomm is positioning its low-power computing experience alongside data center connectivity work, but established suppliers retain extensive ecosystems and customer relationships.
Competition is already vigorous. AMD’s latest data center earnings illustrate the scale of incumbent growth, while AWS’s recent AI usage figures indicate why cloud companies are investing in additional capacity. Those results do not prove that Qualcomm will win share. Its prospects depend on the performance of commercial deployments, customer orders and how its products compare with existing alternatives.
For enterprise technology teams, the immediate implication is to watch compatibility and operating economics, not just headline forecasts. An AI chip offering lower energy consumption is useful only if models run efficiently on its software stack and the hardware can be procured and maintained at scale. The same discipline applies to prospective cloud customers: evaluate availability, latency, memory requirements, security features and all-in cost rather than treating the arrival of a new accelerator vendor as an automatic cost saving.
Higher chipmaking costs remain a near-term risk
Qualcomm warned in its earnings release that the semiconductor industry is facing rising input costs across wafer fabrication, assembly, testing, advanced packaging, memory and other materials. The company is adjusting product pricing and expects the full benefit of those actions to reach QCT gross margins in the second half of fiscal 2027. That timing adds uncertainty as Qualcomm expands into infrastructure products that may have a different margin profile from its traditional chip portfolio.
Its guidance for the first quarter of fiscal 2027 calls for total revenue between $10.4 billion and $11.2 billion, with non-GAAP diluted earnings per share of $2.15 to $2.35. Those projections are not a standalone quarterly data center sales forecast. The company says QCT guidance now includes data center activity, but it has not provided the same level of recurring public segment detail for that business as for handsets, automotive and IoT.
Independent coverage, including The Verge’s account of the earnings call, has focused on the company’s shift toward AI infrastructure as phone demand weakens. Yet investors should distinguish the stated $5 billion fiscal 2027 target from signed customer commitments, realized revenue and profitability. Qualcomm has named strategic collaborations and outlined a roadmap; the quarterly financial statements will be needed to judge the pace of the transition.
The next important evidence will be the company’s fiscal 2027 quarterly results: how quickly new data center projects contribute revenue, whether non-handset growth can offset a weaker smartphone business, and what happens to semiconductor margins as input costs rise. Qualcomm’s November 4 report establishes the scale of the challenge and the ambition. It does not yet establish that the company has completed its move from phone-centric chip designer to major AI infrastructure supplier.
