Qualcomm plans new chip architecture for phones

2026-06-26 · Source: Semafor · Field: Government & Public Sector — Public Policy & Governance, International Relations & Diplomacy, Public Finance & Administration · Depth: Fundamental Awareness, extended

Summary

This daily intelligence brief highlights significant advancements in AI hardware and software, alongside critical geopolitical and economic shifts. Qualcomm plans to integrate its High Bandwidth Compute architecture, initially for data centers, into smartphones by 2028, enabling enhanced on-device AI through vertical chip stacking. OpenAI introduced its custom inference chip, Jalapeño, designed to reduce model serving costs and improve performance per watt, signaling a strategic focus on compute infrastructure. Law firm Cooley launched an AI platform, developed with Legora and trained on Cooley's data, to offer legal assistance to startups for agreements and general inquiries, starting with Y Combinator. Concurrently, the US and Iran signed an interim peace deal, impacting global oil markets and regional stability. China imposed export restrictions on US rare earth companies, escalating trade tensions. The semiconductor market is experiencing surging demand, leading to Apple price increases and increased competition among chip manufacturers.

Key takeaway

For technology executives and investors evaluating future AI infrastructure and market shifts, you should prioritize investments in advanced chip architectures and custom AI silicon, as these are becoming central to competitive advantage and cost efficiency. The convergence of data center chip technology into mobile devices signals a significant shift towards ubiquitous, powerful on-device AI. Additionally, monitor geopolitical developments like rare earth restrictions and trade policies, as they directly impact supply chains and market access for critical components.

Key insights

AI's future hinges on advanced chip architectures and optimized compute infrastructure, amidst shifting global trade and geopolitical landscapes.

Principles

• AI compute infrastructure is a critical competitive battlefront.
• Vertical chip stacking improves data speed and flow.
• AI can be designed to stimulate critical thinking.

Method

Qualcomm's High Bandwidth Compute architecture places memory and compute closer by stacking chips vertically. Cooley's platform uses legal AI trained on proprietary data to analyze documents and answer legal questions.

In practice

• Run more AI models locally on mobile devices.
• Use custom chips to reduce AI model serving costs.
• Utilize AI platforms for startup legal document analysis.

Topics

• Mobile AI
• Custom AI Chips
• Semiconductor Supply Chain
• Legal AI
• Geopolitical Risk
• Quantum Computing Policy

Best for: General Interest, Executive, Investor

Related on AIssential

• AI data center boom is leaving consumer electronics short of chips − even though they don’t use the same kinds — AI data center demand is reorganizing the chip market, creating shortages and higher costs for consumer electronics.
• How the Global AI Race Has Changed — The global AI landscape is rapidly diversifying with new powers and technological fronts emerging beyond the traditional US-China rivalry.
• Qualcomm & The Five Structural AI Inflections — AI infrastructure is reaching a critical density and cost threshold, enabling a new wave of previously impossible products.
• OpenAI Declares the Next Phase of AI — The AI landscape is shifting towards infrastructure-centric development and widespread distribution, marked by major IPOs and novel compute solutions.
• Trump’s China Visit Puts AI Chips and Trade Back in Focus — AI chips are now a strategically vital global resource, akin to oil, influencing geopolitical and economic policies.
• AMD Soars on Blockbuster AI-Fueled Forecast — AI demand is driving significant growth and strategic shifts across semiconductor, cloud, and consumer tech sectors.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Semafor.