Qualcomm (Nasdaq: QCOM): Betting Dragonwing Repeats the Snapdragon Playbook
Qualcomm turned Snapdragon into the default chip inside billions of Android phones by building one flexible platform that dozens of manufacturers could build on top of. It's now trying the identical strategy on robots with Dragonwing, a robotics-specific chip line launched in January 2026 — and its CEO has publicly bet that robotics becomes a larger opportunity than the smartphone business within two years.
Who they are
Qualcomm is best known for Snapdragon, the mobile chip platform that became the default silicon inside most non-Apple smartphones by offering manufacturers a flexible, well-supported reference design they could build products around rather than developing chips from scratch. That playbook — one adaptable platform, broad manufacturer adoption, recurring licensing revenue on top of chip sales — is explicitly what CEO Cristiano Amon has said Qualcomm intends to repeat in robotics.
The company launched Dragonwing at CES 2026, explicitly branded as the "brain of the robot," intended to power everything from industrial autonomous mobile robots up to full-size humanoid robots. The specific model unveiled, the Dragonwing IQ10 Series, was positioned as a robotics reference design in the same spirit as Snapdragon's role in mobile — a chip other companies build robots around rather than compete against.
What they do
Beyond the Dragonwing chip launch, Qualcomm signed a long-term collaboration with Neura Robotics in March 2026 to jointly develop what the companies describe as "brain + nervous system" reference architectures for next-generation humanoid robots — explicitly extending Qualcomm's role beyond selling silicon into co-developing the systems architecture a humanoid needs, a deeper commitment than a typical chip-supply relationship.
CEO Cristiano Amon has been unusually direct in public statements about timeline and scale. At Mobile World Congress in March 2026, he told CNBC that robotics would begin to achieve scale within the next two years and could become a "larger opportunity" for Qualcomm than its historical smartphone business, driven by physical AI breakthroughs reaching commercial maturity by around 2027. Market forecasts he cited put the general-purpose robotics market at roughly $370 billion by 2040, with humanoid robots potentially reaching a $9 trillion market by 2050 — extremely long-dated, aggressive figures that should be read as illustrative context for Qualcomm's strategic rationale rather than near-term guidance.
At Computex 2026, Qualcomm also unveiled Dragonfly, a new data center AI inference chip brand marking its entry into that market, alongside continued Dragonwing robotics platform development — part of a broader "Computing Continuum" strategy to dynamically allocate AI workloads across devices, edge and cloud, of which robotics compute is one specific application.
How they make money
Qualcomm's fiscal third-quarter 2026 results show the company's diversification strategy working, even as its largest segment struggles: total revenue was $9.95 billion, down 4% year over year but ahead of the roughly $9.67 billion analysts expected. QCT handset revenue fell 20% to $5.09 billion amid industry-wide memory supply constraints and softer demand, while QCT Automotive revenue surged 61% to a record $1.59 billion, its 23rd consecutive quarter of double-digit year-over-year growth, and IoT revenue — the category that includes robotics — rose 9% to $1.83 billion.
Guidance for fiscal Q4 2026 called for revenue of $9.7 to $10.5 billion, and management indicated non-handset revenue growth, including data center, would accelerate from 24% in fiscal 2026 to more than 60% in fiscal 2027 — a significant inflection point in Qualcomm's own diversification narrative, of which robotics is one contributing but not solely determining factor.
Where it sits in the value chain
Qualcomm sits in Stage 5 — Compute & robot foundation models, positioned as the low-power edge-compute alternative to NVIDIA's Jetson Thor platform. Where NVIDIA has built its robotics strategy around powerful GPU-based compute paired with its Isaac simulation and GR00T foundation model stack, Qualcomm's pitch centres on power efficiency and cost, the same attributes that made Snapdragon dominant in battery-constrained mobile devices.
The Neura Robotics reference-architecture partnership positions Qualcomm one step closer to the systems-integration layer than a pure chip vendor typically operates, similar in spirit to how Alphabet's Gemini Robotics extends beyond raw compute into the model layer — both companies are betting that owning more of the robot's "brain" architecture, not just supplying a commodity chip, is where durable value accrues.
The bigger trend
Qualcomm's public, specific two-year timeline for robotics reaching commercial scale is one of the more concrete and falsifiable predictions made by any company in this entire chain. That specificity cuts both ways: it gives the market a clear near-term milestone to hold management accountable against, but it also means Qualcomm has more publicly staked its credibility on a fast robotics ramp than most peers, who tend to describe multi-year or decade-long horizons instead.
The Dragonwing strategy's success depends heavily on whether the humanoid and general robotics market actually consolidates around a small number of standard compute platforms the way mobile phones did around Snapdragon and comparable chips, rather than fragmenting across proprietary, vertically integrated silicon the way Tesla is pursuing with its own AI chip designs for Optimus. Which pattern wins will determine whether Qualcomm's platform bet pays off at the scale Amon has publicly projected.
What to watch
- Whether robotics revenue becomes large enough within IoT to warrant separate disclosure, as management has begun explicitly crediting it for growth.
- Progress on the Neura Robotics "brain + nervous system" reference architecture and any resulting commercial products.
- Additional named humanoid or robotics customers adopting Dragonwing, beyond the initial platform launch.
- Whether Amon's two-year timeline for robotics reaching "scale" proves accurate as 2027-2028 approaches.
- Competitive dynamics against NVIDIA's Jetson Thor and any vertically integrated in-house silicon efforts from major robot makers like Tesla.
Related companies
🇺🇸Model-layer counterpart, with Gemini Robotics running on compute platforms including potentially Qualcomm's.
🇰🇷Fleet orchestration software one layer above the chip and model stack Qualcomm and Google supply.
🇺🇸The vertically integrated counterexample, building its own AI chip rather than adopting a platform like Dragonwing — not yet profiled.
FAQ
What is Dragonwing?
Qualcomm's robotics-specific chip platform, launched in January 2026 at CES, designed as a flexible "brain" for robots ranging from industrial autonomous mobile robots to full-size humanoids. It follows the same strategic logic as Snapdragon in mobile phones: one adaptable platform that many robot manufacturers can build on top of.
How much revenue does Qualcomm currently make from robotics?
Robotics revenue is not disclosed as a standalone figure. It falls within Qualcomm's IoT segment, which generated $1.83 billion in fiscal Q3 2026, up 9% year over year, with management specifically crediting industrial, networking and robotics products for that growth, though the robotics-specific portion is not broken out.
What did Qualcomm's CEO say about robotics timing?
Cristiano Amon told CNBC in March 2026 that he expects robotics to begin reaching commercial scale within two years, potentially becoming a "larger opportunity" for Qualcomm than its historical smartphone business. This is a forward-looking management projection, not a guaranteed outcome.
What is the Neura Robotics partnership?
A long-term collaboration announced in March 2026 to jointly develop "brain + nervous system" reference architectures for next-generation humanoid robots, extending Qualcomm's involvement beyond selling chips into co-developing systems-level robot architecture with a named robotics company.