Apple Raises Prices, Micron Rallies | Bloomberg Tech 6/25/2026

Key Concepts

• Memory Supercycle: A structural shift in the semiconductor industry where demand for HBM (High Bandwidth Memory) and standard memory (DRAM/NAND) for AI data centers outstrips supply, leading to long-term price hikes.
• AI Infrastructure Buildout: The massive capital expenditure (CapEx) by hyperscalers to build data centers, which is now transitioning from a pure cost center to generating measurable revenue.
• Agentic AI: AI systems capable of autonomous action and complex workflows, which are driving new hardware requirements and software architectures.
• Custom Silicon (ASICs): Specialized chips designed for specific AI workloads, allowing companies like Qualcomm to bypass traditional bottlenecks like HBM.
• Disaggregated Compute: The trend in data centers moving away from monolithic GPU reliance toward a mix of specialized CPUs, accelerators, and custom memory solutions.

---

1. The Memory Crunch and Hardware Pricing

Micron’s recent earnings report shattered Wall Street expectations, with the company forecasting up to $51 billion in revenue for the fourth quarter. Management indicated that AI-driven memory shortages could persist beyond 2027.

• Apple’s Response: Apple has implemented "unprecedented" price hikes across its product lineup (iPad, Mac, HomePod, Apple TV, Vision Pro). Analysts note this is a direct consequence of the sustained high cost of memory and storage components.
• Market Impact: Carol Schleif (BMO Wealth Management) notes that tech is no longer a purely deflationary force. The economy is undergoing a "regime change" where massive infrastructure investment is creating endemic core inflation in the short term.

2. Qualcomm’s Strategic Pivot to Data Centers

Qualcomm CEO Cristiano Amon outlined an ambitious growth strategy during their Investor Day, targeting $15 billion in annual AI data center revenue by fiscal 2029 and $40 billion in non-smartphone business.

• Differentiation Strategy: Qualcomm is focusing on power efficiency (leveraging their mobile battery-conscious design heritage) and custom ASICs.
• Technical Innovation: Amon highlighted a new memory solution designed by Qualcomm that avoids the need for expensive HBM, reducing thermal output and power consumption.
• Partnerships: Qualcomm confirmed a two-generation CPU contract with Meta, emphasizing that their existing consumer-side relationship (Snapdragon/Quest) was instrumental in securing data center wins.

3. Software Evolution: Figma’s "Intelligent Canvas"

Figma unveiled a platform overhaul designed for the "AI era," moving beyond static design to a full-stack digital creation environment.

• Key Features: The platform now includes "code layers" for interactive design, generative effects, and "Figma Motion."
• Human-in-the-Loop: CEO Dylan Field argued that AI tends to produce "average" outputs. He posits that human designers are the "differentiators" who use AI to raise the ceiling of creativity rather than replace it.
• Modular AI: Figma is model-agnostic, allowing users to swap AI models to maintain control over their proprietary brand assets and "ground truth."

4. The AI Revenue Tipping Point

Azim Azar (Exponential View) presented research showing that the AI buildout is beginning to pay off.

• Data: Global AI sales (excluding China) reached $25 billion in Q1 2026, surpassing the $21 billion in depreciation from hardware investments.
• Growth Rate: The AI economy is growing at an annualized rate of approximately 200%, which Azar notes is roughly three times faster than the growth rates of the internet or mobile computing eras.

5. Regulatory and Talent Risks

• Public Sentiment: Wall Street strategists are warning that public dissatisfaction regarding job displacement and energy consumption could lead to anti-AI policies, such as moratoriums on data center construction or new taxes on AI producers.
• Talent Drain: Google is facing a wave of high-profile departures (e.g., Jonas Adler, Alexander Pritzell, Noam Shazeer) to competitors like Anthropic and OpenAI. Reports suggest that internal competition for "precious" computing resources has become a point of friction, contributing to the exodus of top-tier AI researchers.

---

Synthesis

The tech industry is currently defined by a "bottleneck" in memory supply and a massive, rapid rewiring of the global economy for AI. While hardware costs are forcing consumer electronics price hikes, the "AI buildout" is finally crossing the threshold into profitability. Companies like Qualcomm are successfully diversifying by targeting the data center with power-efficient, custom silicon, while software platforms like Figma are evolving to integrate AI agents into professional workflows. However, the sector faces significant headwinds, including potential regulatory intervention driven by public sentiment and the ongoing challenge of retaining top-tier AI talent.