Qualcomm Bets Big on AI to Offset Smartphone Slump

Qualcomm is injecting a massive dose of artificial intelligence into its chip portfolio, hoping that on-device AI capabilities can compensate for the structural weaknesses plaguing the global smartphone market. The question investors, analysts, and competitors are all asking: is this AI 'adrenaline shot' powerful enough to cure what ails the mobile giant?

The San Diego-based chipmaker has spent the last 2 years aggressively repositioning itself as an AI-first semiconductor company, pushing its Snapdragon platform as the go-to solution for running large language models and generative AI workloads directly on smartphones, PCs, and automobiles — without relying on the cloud.

Key Takeaways

• Qualcomm's smartphone revenue still accounts for roughly 60% of total sales, but growth has stagnated amid a global handset slowdown
• The company's Snapdragon 8 Elite chip can run AI models with up to 13 billion parameters on-device
• Qualcomm is expanding beyond phones into PCs, automotive, and IoT — all with an AI-centric pitch
• On-device AI is positioned as a key differentiator against rivals like MediaTek, Apple, and Samsung Exynos
• The AI PC market, where Qualcomm competes with its Snapdragon X Elite chips, is projected to reach $200 billion by 2028
• Qualcomm's automotive design-win pipeline has surpassed $45 billion, largely driven by AI-powered digital cockpit solutions

The Smartphone Market's Structural Problem

The global smartphone market shipped approximately 1.2 billion units in 2024, a modest recovery from the post-pandemic downturn but still well below the 1.4 billion peak seen in 2017. Replacement cycles have stretched to over 4 years in many markets, and consumers increasingly see little reason to upgrade when year-over-year improvements feel incremental.

For Qualcomm, this is an existential challenge. The company's QCT (Qualcomm CDMA Technologies) segment — which sells Snapdragon chips to smartphone OEMs like Samsung, Xiaomi, OnePlus, and others — remains the revenue engine. But when fewer phones sell, or when they sell at lower price points, Qualcomm's top line suffers directly.

Making matters worse, Qualcomm faces an ongoing threat from its largest customer. Samsung continues developing its own Exynos processors, and Apple has never used Qualcomm's application processors. The modem business, once a guaranteed revenue stream, also faces long-term risk as Apple works to bring its own 5G modem in-house — a transition already underway with the iPhone 16e.

On-Device AI: Qualcomm's Strategic Counterattack

Qualcomm's answer to these headwinds is straightforward: make AI the reason people upgrade their phones. The company argues that generative AI workloads — real-time language translation, AI photography, intelligent assistants, on-device content generation — require significantly more processing power than anything current phones deliver.

The Snapdragon 8 Elite, launched in late 2024, features a dedicated Hexagon NPU (Neural Processing Unit) capable of delivering over 75 TOPS (trillion operations per second). This is a dramatic leap from just 2 years prior, when the Snapdragon 8 Gen 2 offered roughly 15 TOPS.

Key on-device AI capabilities Qualcomm is pushing include:

• Running LLMs up to 13 billion parameters locally, enabling private and low-latency AI assistants
• Real-time multimodal AI processing — combining text, image, and voice inputs simultaneously
• AI-enhanced photography and video with computational RAW processing and generative fill
• On-device speech-to-speech translation with sub-second latency
• Personalized AI agents that learn user behavior without sending data to the cloud
• Integration with frameworks like Meta's Llama, Microsoft Phi, and Google Gemini Nano

The pitch to consumers is compelling: faster, smarter, and more private AI experiences that do not depend on internet connectivity or cloud subscriptions.

Can AI Actually Drive Smartphone Upgrades?

This is the $100 billion question. Historically, few single features have proven capable of triggering mass smartphone upgrade cycles. The last major catalyst was 5G, which drove a wave of replacements between 2020 and 2022 but quickly lost its pull once networks matured and coverage expanded.

AI could be different — but only if the use cases are genuinely transformative. Early signs are mixed. Samsung's Galaxy AI features, powered partly by Snapdragon silicon, have been well-received in reviews but have not yet driven a measurable surge in upgrade intent. Apple's Apple Intelligence, while generating enormous buzz, has been criticized for slow rollout and limited functionality at launch.

The challenge for Qualcomm is that it does not control the software experience. It provides the hardware foundation, but the actual AI applications that consumers interact with are built by Google, Samsung, and third-party developers. If those software experiences fail to impress, even the most powerful NPU becomes irrelevant to the average buyer.

Moreover, there is a competitive dynamic at play. MediaTek's Dimensity 9400, targeting the premium Android segment, offers comparable AI performance at a lower price point. As MediaTek gains share among Chinese OEMs — who collectively account for over 50% of global Android shipments — Qualcomm's pricing power in mobile comes under pressure.

Beyond Phones: Diversification as the Real AI Play

Perhaps the more important story is not whether AI saves Qualcomm's smartphone business, but whether AI accelerates the company's diversification into higher-growth markets.

Automotive is the standout. Qualcomm's Snapdragon Digital Chassis platform powers infotainment, advanced driver-assistance systems (ADAS), and connected car features for automakers including BMW, Mercedes-Benz, General Motors, and Hyundai. The company's automotive design-win pipeline exceeds $45 billion — a figure that has roughly doubled in 2 years.

PCs represent another frontier. The Snapdragon X Elite and Snapdragon X Plus processors, based on ARM architecture, have brought Qualcomm into direct competition with Intel and AMD in the laptop market. Microsoft's Copilot+ PC initiative, which requires a minimum of 40 TOPS of NPU performance, has effectively made Qualcomm the launch partner for the AI PC era.

Early results show promise. Devices from Dell, HP, Lenovo, and Microsoft Surface running Snapdragon X chips have earned praise for battery life and AI performance, though x86 app compatibility via emulation remains a concern for some enterprise users.

IoT and Industrial applications round out the portfolio. Qualcomm's AI-capable chips are finding their way into smart cameras, robotics, edge computing gateways, and retail analytics systems — markets less glamorous than smartphones but growing at double-digit rates.

Industry Context: The On-Device AI Arms Race

Qualcomm is not alone in recognizing on-device AI as a strategic battleground. The entire semiconductor industry is pivoting toward NPU-centric architectures.

Apple's A18 Pro and upcoming M5 chips continue to set benchmarks for on-device ML performance. Google's Tensor G4 prioritizes AI workloads for its Pixel lineup. Intel and AMD are embedding NPUs into every new laptop processor. Even NVIDIA, the undisputed king of cloud AI, is eyeing edge and on-device inference through its Jetson and automotive platforms.

What distinguishes Qualcomm is its breadth. No other company offers AI-optimized silicon across smartphones, PCs, automobiles, XR headsets, and IoT simultaneously. This cross-platform reach gives Qualcomm a unique ability to offer developers a unified AI stack — write once, deploy everywhere — that could become a powerful ecosystem moat over time.

The company's AI Hub, launched in 2024, now supports over 100 optimized AI models that developers can deploy across Snapdragon platforms. This developer ecosystem play mirrors what Apple achieved with its silicon transition and could prove just as strategically important.

What This Means for Developers and Businesses

For developers building AI-powered applications, Qualcomm's push creates tangible opportunities:

• Lower latency: On-device inference eliminates round-trip cloud delays, enabling real-time AI features
• Cost savings: Running AI locally reduces cloud compute costs, particularly for high-volume consumer apps
• Privacy advantages: Sensitive data never leaves the device, simplifying compliance with GDPR, HIPAA, and other regulations
• Offline capability: AI features work without internet connectivity, critical for automotive, industrial, and emerging market use cases

For businesses evaluating AI strategies, the key insight is that the 'AI everywhere' vision is becoming hardware-ready faster than many expected. Edge AI is no longer a future concept — it is shipping in hundreds of millions of devices today.

Looking Ahead: Will the AI Bet Pay Off?

Qualcomm's next major milestone arrives with the Snapdragon 8 Elite 2 (expected late 2025), which is rumored to push NPU performance beyond 100 TOPS and support models with 20+ billion parameters on-device. If realized, this would bring smartphone AI capabilities closer to what currently requires a cloud GPU.

The company's fiscal trajectory will reveal whether AI is genuinely moving the needle. Wall Street expects Qualcomm's QCT handset revenue to grow in the mid-single digits in fiscal 2025 — respectable but hardly transformative. The real upside lies in automotive (projected 30%+ growth) and IoT/PC segments.

Ultimately, Qualcomm's AI 'adrenaline shot' may not fully cure the smartphone market's chronic ailments — elongating upgrade cycles and intensifying competition are structural forces that no single technology can reverse. But AI does give Qualcomm a credible narrative for premium pricing, a path to diversification, and a platform strategy that extends well beyond the phone in your pocket.

The question is no longer whether Qualcomm can build great AI hardware. It clearly can. The question is whether the rest of the ecosystem — app developers, OEMs, and ultimately consumers — will deliver the software experiences that make that hardware matter.