• June 26, 2026

Yesterday I appeared on Yahoo Finance to discuss the information coming out of the investor day that Qualcomm held in New York City this week. We also discussed Micron’s great financial numbers and how sustainable that performance may be. You can watch an edited excerpt of my in-studio conversation with Julie Hyman via the link below, or read on for an AI-generated, human-edited overview and summary of the much longer complete discussion.

Link to edited segment.

Overview

Qualcomm is expanding into the datacenter and AI chip markets to capture a share of a growing trillion-dollar industry. The company is introducing High-Bandwidth Compute, a stacking technology designed to improve memory bandwidth and energy efficiency without requiring TSMC’s advanced chip packaging. Qualcomm is actively diversifying its revenue streams, with its projections showing that smartphone chips will comprise only a third of its business by 2029, bolstered by heavy growth in the automotive sector and elsewhere. Meanwhile, Micron has achieved massive revenue beats, driven by price increases rather than volume gains, while facing potential future competition from Chinese manufacturers.

Outline

Qualcomm’s New Architecture and Datacenter Ambitions

• Julie Hyman of Yahoo Finance introduces Patrick Moorhead to discuss Qualcomm’s recent investor and analyst event.
• Moorhead explains that Qualcomm is entering the trillion-dollar datacenter market in a time when customers are facing growing energy constraints and demanding new architectures.
• He highlights Qualcomm’s new High-Bandwidth Compute (HBC) stacking technology, which places compute directly underneath memory on a chip.
• Moorhead notes that this technology improves efficiency and lowers costs by eliminating the need for TSMC’s advanced CoWoS packaging.
• Moorhead also mentions that Qualcomm already has major customers for this technology, including rumored custom solutions for Amazon and ByteDance.

AI-Aided Chip Design and Qualcomm’s Revenue Diversification

• Hyman asks whether advanced AI will eventually be able to design more efficient and cost-effective chips on its own.
• Moorhead confirms that major EDA players like Synopsis and Cadence, as well as chip companies themselves, are already using AI in their design pipelines.
• He points out that while AI can design chips, it cannot solve the physical manufacturing bottlenecks and energy constraints faced by the industry today.
• Hyman notes that investors have been waiting for Qualcomm to reduce its heavy revenue reliance on smartphone chips.
• Moorhead states that Qualcomm has demonstrated real diversification by tracking to become the largest automotive chip manufacturer within two years (according to Qualcomm itself).
• He shares a company estimate that smartphones will make up only a third of Qualcomm’s business by 2029.

Micron Revenue Beats and Pricing Strategy

• Hyman shifts the conversation to Micron, comparing its recent massive financial beats to historic performance overperformance seen by NVIDIA.
• Moorhead emphasizes that memory is a critical strategic gatekeeper for AI advancement, but notes Micron’s revenue growth has come from price increases rather than higher unit volume.
• Hyman counters that many of Micron’s customers are locked into five-year long-term agreements for these prices.
• Moorhead agrees, but notes that these long-term agreements are often fungible in terms of price and that Micron has previously renegotiated pricing within these contracts.

Micron’s Market Risks and NVIDIA Comparison

• Moorhead warns that Micron’s pricing power could face risks if Chinese memory manufacturers like CXMT bring significant supply online.
• Hyman asks what it means for companies like Lenovo, Dell, and HP to “qualify” this alternative Chinese memory.
• Moorhead defines qualification as the rigorous quality and performance testing process that serves as the precursor to major purchase orders.
• Moorhead also clarifies that Micron’s situation contrasts with NVIDIA’s, whose massive growth over five years was driven by selling more physical silicon rather than raising prices on the same chip density.

Companies Mentioned

• Qualcomm
• TSMC
• Amazon
• ByteDance
• Synopsis
• Cadence
• NXP
• Renesas
• Apple
• Micron
• Lenovo
• Dell
• HP
• NVIDIA
• CXMT

Keyword Summary

Qualcomm, HBC, HBM, memory chips, automotive chips, Micron, semiconductors

Recent Posts