Qualcomm is transforming itself from a mobile-focused chipmaker into a broad-based AI infrastructure supplier by embedding neural processing units across smartphones, PCs, vehicles, and industrial devices. The company's Hexagon NPU, combined with its CPUs and GPUs, enables generative AI models to run directly on edge devices without relying on cloud servers. This shift reflects a fundamental change in how the semiconductor industry is approaching artificial intelligence deployment.

What Are Neural Processing Units and Why Do They Matter?

Neural processing units, or NPUs, are specialized chips designed specifically to accelerate artificial intelligence workloads on individual devices. Unlike general-purpose processors, NPUs handle the mathematical operations required for machine learning inference with exceptional efficiency. Qualcomm's Hexagon NPU is integrated into its Snapdragon processors, which power everything from flagship smartphones to edge computing systems. This architecture allows devices to process AI tasks locally, reducing latency and improving privacy by keeping sensitive data off cloud servers.

The market opportunity is substantial. Global spending on artificial intelligence infrastructure is projected to rise at an annual rate of nearly 30 percent, climbing from approximately $235 billion in 2024 to roughly $630 billion through 2028. While much of this investment has historically flowed toward data center GPUs, the emergence of capable on-device AI is creating a parallel growth vector for companies like Qualcomm.

How Is Qualcomm Expanding Beyond Smartphones?

Qualcomm's strategy centers on leveraging its deep intellectual property in cellular connectivity, low-power computing, and chip integration to become a major provider of on-device and edge AI technology. The company's most recent investor day presentation in June outlined an ambitious diversification plan that dramatically reduces its dependence on the mature smartphone market. Management raised its fiscal 2029 target for non-handset revenue to $40 billion and projected more than $15 billion in data-center revenue, $10 billion in automotive revenue, and more than $14 billion from the internet of things, or IoT.

This expansion strategy addresses a critical market reality: handsets are expected to represent only about one-third of Qualcomm's semiconductor revenue by 2029. The company is positioning its Snapdragon processors and Hexagon NPU technology across multiple verticals:

• Automotive: AI-powered features in vehicles require real-time processing without cloud connectivity, making on-device NPUs essential for autonomous driving, voice assistants, and predictive maintenance systems.
• Industrial IoT: Manufacturing equipment, robotics, and smart sensors need low-latency AI inference at the edge, where cloud connectivity may be unreliable or impractical.
• Data Center Infrastructure: Qualcomm is now competing in the server market with processors designed to handle AI workloads alongside traditional computing tasks.
• Wearables and Robotics: Devices with limited power budgets benefit from NPU-accelerated AI that reduces battery drain compared to general-purpose processors.

According to Mark Vena, CEO of SmartTech Research, this diversification is a significant competitive advantage. Vena noted that "Qualcomm's broadening presence in automotive, industrial IoT, AI PCs and now AI data center infrastructure provides investors many growth engines rather than just one market".

Vena

What Does This Mean for the Broader Semiconductor Industry?

Qualcomm's pivot reflects a larger industry trend: the shift from centralized cloud AI to distributed, on-device intelligence. While hyperscalers like Alphabet, Amazon, Meta Platforms, and Microsoft continue massive capital expenditures on data center infrastructure, the emergence of capable edge AI is creating new demand for specialized chips. Qualcomm's Hexagon NPU represents one approach to this challenge, but it is not alone. Other chipmakers are developing competing neural engines and integrated AI systems.

The competitive landscape is intensifying. Nvidia remains the dominant player in AI accelerators, but its focus on data center GPUs leaves room for companies like Qualcomm to capture the on-device and edge segments. Qualcomm's recent quarterly results were modest, reflecting the maturity and cyclicality of the smartphone market. However, management's long-term guidance suggests confidence in the company's ability to grow beyond its traditional handset business.

Semiconductor investors should expect continued volatility as the industry navigates this transition. According to Vena, "Investors should expect more volatility as semiconductor stocks have surged sharply and expectations are sky high. But in the longer term, fundamentals remain compelling and valuation discipline and execution are much more important as AI spending broadens out beyond the first wave of infrastructure investment".

Why On-Device AI Is Becoming Critical for Enterprises

The shift toward on-device AI processing addresses several practical challenges for businesses and consumers. Running AI models locally reduces dependency on cloud connectivity, improves response times, and enhances data privacy by keeping sensitive information on the device. For enterprises deploying AI across distributed networks of devices, this approach reduces bandwidth costs and eliminates latency issues associated with cloud round-trips. Qualcomm's Hexagon NPU is engineered to handle these workloads efficiently while consuming minimal power, a critical constraint for mobile and wearable devices.

The company's positioning in automotive and industrial IoT markets is particularly strategic. Autonomous vehicles require real-time AI inference for object detection, path planning, and decision-making. Industrial robots and manufacturing systems need AI-powered quality control and predictive maintenance without relying on cloud connectivity. Qualcomm's integrated approach, combining CPUs, GPUs, and NPUs on a single chip, simplifies system design and reduces power consumption compared to discrete components.

As the semiconductor industry continues to evolve, Qualcomm's neural processing unit strategy represents a bet that the future of AI is distributed, not centralized. Whether this bet pays off will depend on the company's ability to execute across multiple markets and compete against both established players and new entrants focused on edge AI.