Bloomberg Tech
2026-06-26 · Hosted by Caroline Hyde, Ed Ludlow · Bloomberg / iHeartMedia
Executive Summary
Memory dominated the show as Micron shattered Wall Street expectations, forecasting up to $51 billion in fiscal Q4 revenue (vs. ~$43B street estimate) and warning that AI-driven memory shortages could persist beyond 2027, keeping prices structurally high. The memory crunch forced Apple into largely unprecedented price hikes across the iPad, Mac, HomePod, Apple TV and Vision Pro, with executives hinting iPhone increases could follow in September. Qualcomm CEO Cristiano Amon detailed an aggressive data-center push targeting $5B in fiscal 2027 and $15B by fiscal 2029, while a new exponential-view analysis pegged annualized AI revenue at $170B (growing ~200% YoY). The episode also flagged a rising political/PR backlash against AI and a wave of researcher departures from Google to Anthropic and OpenAI.
Key Stories & Changes
1. Micron’s Blowout Forecast Resets the Memory Market
Micron forecast as much as $51 billion in fiscal Q4 revenue, well above all street estimates (~$43B)
Management says AI-driven memory shortages could last beyond 2027; “no line of sight to supply improving”
Stock up ~12% on the session (had been higher at the open); priced around $1,190 a share
Bloomberg Intelligence’s Jake Silverman: contracts now extending out to 2030, a structural break from historical monthly/quarterly contracts
DRAM and NAND prices estimated up 200–300% this year
2. Apple Raises Hardware Prices on Memory Crunch
Per Mark Gurman, price hikes are “largely unprecedented in Apple’s modern history,” spanning iPad, Mac, HomePod, Apple TV, Vision Pro
Excludes iPhone, AirPods, Apple Watch — for now; Apple “seriously hinted” at more increases (potentially iPhone/Watch in September)
Apple explicitly tied the move to AI data center memory demand, saying it held off “as long as possible”
3. Qualcomm Investor Day & CEO Interview
Projecting $15B+ in annual AI data-center revenue by fiscal 2029 and $40B in non-smartphone sales (double the prior two-year target)
Near-term $5B data-center target for fiscal 2027 called a “high-confidence number” with one large US hyperscaler and one large China hyperscaler customer
Differentiation via power efficiency, a non-HBM memory solution, ASICs (Alphawave), and a CPU contract with Meta for two generations
Acquisitions cited: Alphawave, Ventana, Nuvia, modular (Chris Lattner), Arduino; future M&A focused on data center and robotics
Sees the China Android market roughly flat to down in 2027 — a memory-supply issue, not demand
4. AI Revenue Reaches a Tipping Point
Azeem Azhar (exponential view): global AI sales ex-China hit $25B in Q1 2026, exceeding expected data-center/chip depreciation of $21B
Annualized run-rate now ~$170B, growing ~200% YoY; Anthropic’s tear kept growth from decelerating
AI real revenue growing ~3x faster than the internet and mobile did
5. Anti-AI Political Backlash as a New Tech Risk
Bloomberg’s Matthew Griffin: strategists warn public anger (jobs, electricity prices) could spur anti-AI policy
Worst cases: a national moratorium on data center construction or a tax targeting AI producers
Cities/counties have paused construction; a moratorium sits on the New York governor’s desk; Ohio and Illinois paused data-center tax breaks
6. Google AI Talent Exodus
Senior researchers Jonas Adler and Alexander Pritzel (key Gemini contributors) leaving Google for Anthropic
Follows Noam Shazeer (to OpenAI) and Nobel laureate John Jumper (to Anthropic)
Bloomberg’s Julia Love: compute for one Shazeer project was reportedly shifted to a larger London pre-training team before his exit
7. Figma’s AI-Era Platform Overhaul
CEO Dylan Field unveiled an “intelligent canvas” with code layers, AI agent plugins, generative shaders, and Figma Motion
Model-agnostic/modular approach; “Figma Weave” composes models and workflows accessible via API
Trends Identified
1. Memory as a Structural, Not Cyclical, Constraint
Micron’s commentary and the move to multi-year contracts out to 2030 mark a departure from memory’s historically cyclical, deflationary pattern. With no line of sight to supply relief and prices up 200–300% this year, the entire device chain — from Apple to PC makers — must now plan around persistently high component costs.
2. Tech Turns Inflationary
BMO’s Carol Schleif framed an “economic regime change”: tech, long a reliable deflationary force, is now pushing prices higher as the economy rewires for AI after decades of infrastructure underinvestment. The interim period implies higher endemic core inflation until late-decade productivity gains materialize.
3. The AI Build-Out Finally Shows Revenue
Exponential view’s data suggests the demand side of AI is no longer a black box — quarterly revenue now exceeds hardware depreciation, with annualized run-rate near $170B and ~200% growth, reframing the “where’s the return” debate.
4. Diversification Beyond Smartphones
Qualcomm’s aggressive data-center, automotive, industrial and robotics targets exemplify how chip incumbents are repositioning around agentic AI and disaggregated compute rather than today’s GPU-centric architecture. —-
Sentiment Analysis
Overall Market Sentiment: Bullish but Cautious
The dominant mood was strong AI demand validation (Micron, Qualcomm, AI revenue), tempered by inflation, political backlash, and talent-war anxieties.
Risk Factors Highlighted
Memory supply shortage persistence: Shortages potentially beyond 2027 keep input costs high across the device chain.
Apple margin/demand pressure: Unprecedented price hikes risk demand destruction and margin erosion, with iPhone potentially next.
Tech-driven inflation: Higher endemic core inflation as the economy rewires for AI.
Anti-AI policy intervention: Data-center construction moratoriums or AI taxes could slow hyperscaler profit growth.
AI talent flight: Departures from Google to Anthropic/OpenAI threaten Gemini’s competitive position.
Hyperscaler ROI uncertainty: Markets still question returns on the massive capex deployed.
Qualcomm execution/obsolescence risk: New data-center entrant facing annual-cadence competition from Nvidia/AMD.
Geopolitical/export controls: China customer exposure governed by evolving US export rules.
This episode was covered in today’s The Market Signal — 2026-06-26, a cross-source synthesis of multiple podcast reports.