At its high-stakes Investor Day 2026 in New York City, U.S. wireless chip giant Qualcomm Incorporated unveiled a massive structural transformation, forecasting that its newly formed data center division will generate more than $15 billion in annual revenue by fiscal year 2029.

The announcement triggered a massive wave of Wall Street optimism, sending Qualcomm shares surging over 12% to 15% in after-hours trading. The explosive rally added tens of billions to Qualcomm’s market capitalization as investors cheered the company’s aggressive and credible blueprint to break into the booming AI cloud infrastructure market, historically dominated by Nvidia.

1. The 2029 Diversification Roadmap

The $15 billion data center target sits at the heart of an aggressively updated long-term guidance framework presented by CEO Cristiano Amon and CFO Akash Palkhiwala. Qualcomm is actively moving to decouple its financial future from the cyclical smartphone market:

  • The Non-Handset Leap: Qualcomm nearly doubled its fiscal 2029 non-handset chip revenue target to $40 billion, up from its previous projection of $22 billion.
  • Inverting the Smartphone Core: By 2029, traditional smartphone handset chips—previously Qualcomm’s absolute stronghold—are projected to shrink to only about one-third of the company’s total chip business (QCT) revenues.
  • Near-Term Benchmarks: The data center division is not a far-off dream; management outlined a clear checkpoint, projecting the business to bring in $5 billion by fiscal 2027, with $1 billion of that pool already locked in from customized silicon clients.
                  [ Qualcomm QCT Revenue Transformation — FY2029 ]

  Old Outlook ──► Handset Monolith (Nearly 70-80% of revenue)
  
  New Outlook ──► Handsets (~33%) ──► Non-Handset Matrix (~67% / $40 Billion)
                                        │
                    ┌───────────────────┼───────────────────┐
                    ▼                   ▼                   ▼
           [Data Center: $15B+] [IoT/Robotics: $14B+] [Automotive: $10B]

2. Anchoring with Big Tech: Meta and Microsoft Sign On

To prove that it isn’t arrived late to the cloud party, Qualcomm paired its long-term financial modeling with immediate, premier enterprise customer validations:

The Meta Partnership (Dragonfly C1000)

Qualcomm officially named Meta Platforms as a major launch customer for its data center chips. Meta has entered a multi-generational supply agreement to deploy Qualcomm’s brand-new Dragonfly C1000 data center CPUs to power its massive, expanding server fleets starting in the second half of 2028. Mark Zuckerberg noted that the chips will play a core role in building out Meta’s infrastructure for generalized AI.

The Microsoft Validation & Custom Hyperscalers

Microsoft CEO Satya Nadella also expressed active backing, stating the company looks forward to building out cloud capacity leveraging Qualcomm’s data center hardware. Furthermore, data center chief Tony Pialis revealed that Qualcomm has secured two additional, unnamed global hyperscaler clients for custom chip-design services, with initial revenues set to hit the balance sheet before the end of this calendar year.

3. The Technical Disruption: “High Bandwidth Compute” (HBC)

What caught the attention of hardware analysts is Qualcomm’s distinct architectural strategy. Rather than competing head-to-head on Nvidia’s terms, Qualcomm is pioneering a category it calls High Bandwidth Compute (HBC):

  • Bypassing the HBM Bottleneck: Standard high-end AI servers are constrained by an expensive shortage of High Bandwidth Memory (HBM) chips. Qualcomm’s architecture is engineered to run high-volume AI inference workloads using significantly cheaper, highly dense memory chips typically optimized for premium smartphones and laptops.
  • The Cost-Per-Performance Advantage: By leveraging these components, Qualcomm claims its upcoming AI250 accelerator (sampling in mid-2027 with a massive 133 TB/s of effective memory bandwidth) will provide cloud builders with an unparalleled performance-to-cost ratio, drastically lowering the power and capital expenditure required to run live AI models.

4. Buying Software Moats: The $3.9 Billion Modular Deal

Recognizing that hardware is only as good as the software ecosystem supporting it, Qualcomm announced a definitive agreement to acquire AI software infrastructure standout Modular for approximately $3.9 billion in an all-stock transaction.

                               ┌──► Hardware Strategy: Low-power, low-cost HBC inference chips
                               │
[Qualcomm's Data Center Stack] ┼──► Software Layer: $3.9B Modular acquisition (Unified MAX/Mojo execution)
                               │
                               └──► Initial Scale: Target $5 Billion segment run-rate by FY2027

Modular is the creator of the MAX engine and Mojo programming language, which allow developers to seamlessly train and run complex AI models across fragmented architectures (CPUs, GPUs, and NPUs) without having to rewrite underlying code bases for specific vendor environments.

Integrating Modular directly into Qualcomm’s pipeline gives the chipmaker an immediate, developer-friendly software stack, positioning it as a highly integrated platform company capable of targeting a projected $1.7 trillion total addressable market across edge, automotive, and cloud computing by 2030.