AI Is Probably Underestimated, Qualcomm CEO Says
By Bloomberg Technology
Key Concepts
- Qualcomm's Diversification: Moving beyond traditional mobile handsets into automotive, data center, and mixed reality.
- Premium Tier Expansion: Growth in higher-end smartphones, particularly in emerging markets like India and China, benefiting Qualcomm.
- Snapdragon Digital Chassis: Qualcomm's integrated digital platform for the automotive industry.
- Personal AI Devices: New category of wearable devices (e.g., smart glasses) leveraging AI capabilities.
- AI Inference: The process of deploying AI models at scale in data centers, a key focus for Qualcomm.
- Power Efficiency in Data Centers: Qualcomm's emphasis on developing power-efficient architectures for AI inference, stemming from their mobile device DNA.
- Compute Density: Qualcomm's goal of maximizing processing power within a given space in data centers.
- Market Timing and Innovation: The inherent uncertainty in predicting the exact timing of market trends, but a clear underlying trend of increasing compute demand.
Qualcomm's Performance and Strategy
Qualcomm expresses strong confidence in its current execution and strategy, stating that "everything we said we're going to do, we're actually doing better." The company highlights its expansion beyond handsets into various new markets, noting its ability to produce chips ranging from 5 watts for earbuds to 500 watts for data centers. This broad capability is seen as a significant opportunity.
Smartphone Business and Android Expansion
The smartphone market, described as the "bread and butter" of Qualcomm's business, is experiencing a significant trend: the expansion of the premium tier. While overall handset market units may not have returned to pre-COVID levels, the increasing adoption of premium devices (like iPhones and Galaxies) globally, including in the United States, China, and India, is driving growth for Qualcomm. This trend, coupled with phones becoming more capable with increased compute and AI features, is boosting silicon content and contributing to the growth of Android with Snapdragon. The company states that "the upside that we have seen above expectations is really Android expansion for Qualcomm."
Diversification Beyond Mobile
Qualcomm is actively diversifying away from its traditional mobile focus and reducing reliance on Apple. Key growth areas include:
- Automotive: This segment is identified as the "fastest moving" for Qualcomm. The company is experiencing significant year-over-year growth by gaining market share as cars evolve into "computer on wheels." Qualcomm provides a digital platform for the automotive industry called the Snapdragon Digital Chassis. Non-Apple related growth in the company was approximately 18% in the last fiscal year, with automotive segments performing significantly higher.
- Personal AI Devices: Qualcomm is optimistic about new categories of mobile devices, termed "personal AI devices," such as smart glasses. The company anticipates significant revenue from this segment, with projections of $2 billion by fiscal year 2029, and notes they are "way ahead of that" in terms of current performance, especially with partners like Meta.
Qualcomm's Approach to Data Center AI Inference
Qualcomm is positioning itself as a strong contender in the data center AI market, particularly focusing on inference. The company acknowledges the significant investment and competition in this space from players like NVIDIA, AMD, Broadcom, Marvell, and numerous startups.
Qualcomm's differentiator lies in its technological competence and its focus on developing an architecture optimized for inference. The company emphasizes its strength in patent applications, being the "number one company in America in terms of patent applications" last year.
Key aspects of Qualcomm's data center AI strategy:
- Focus on Inference: Recognizing that profitability in the AI data center market hinges on successful and scalable inference, Qualcomm is developing solutions for this critical stage.
- Power Efficiency: Drawing from its mobile DNA, Qualcomm is prioritizing power efficiency in its data center architectures. This is crucial given the high energy demands of AI workloads.
- Optimized Architecture: Qualcomm is developing an architecture specifically designed for inference, aiming for the "highest possible compute density" with "very efficient power." This is positioned as a potential "post GPU" architecture.
- Market Opportunity: Qualcomm believes the AI market is so large that even a small share will translate into "multiple of billions of dollars" in revenue.
Market Projections and Concerns
The discussion touches upon the ambitious projections for data center investment, with Morgan Stanley estimating $3 trillion by 2028. While acknowledging the potential for such growth, the conversation also explores potential risks:
- Power Constraints: A concern is raised about whether power limitations could hinder the realization of these massive data center demands, potentially leading to a need for innovation driven by necessity.
- Timing Uncertainty: While the underlying trend of increasing compute demand is clear across phones, cars, and PCs, the exact timing of market shifts and the realization of these projections remains a subject of debate.
Qualcomm's perspective is that the "need for compute is very clear" and that AI will drive innovation and competition. They liken the current situation to the early days of the internet, where initial projections, while significant, ultimately underestimated the eventual scale and impact. Qualcomm is preparing for this future competition and believes that "traffic on the long run air is probably underestimated, is going to be bigger."
Conclusion
Qualcomm is demonstrating strong execution and strategic foresight, successfully diversifying beyond its core mobile business into high-growth areas like automotive and AI. The company's focus on premium smartphone segments, its innovative Snapdragon Digital Chassis, and its targeted approach to AI inference in data centers, emphasizing power efficiency and compute density, position it for continued success. While market timing remains a variable, Qualcomm is confident in the fundamental and growing demand for compute power, viewing it as a significant opportunity for innovation and market leadership.
Chat with this Video
AI-PoweredHi! I can answer questions about this video "AI Is Probably Underestimated, Qualcomm CEO Says". What would you like to know?