Semiconductors

Semiconductors Insights

Research notes on earnings, market structure, and financial statements. Each piece starts with the conclusion, then lays out the evidence and implications.

2026-06-24

Abstract data center roadmap graphic showing Qualcomm moving from devices to cloud

Strategy

Jun 24, 2026

QCOM•10 min read

Qualcomm Data Center Pivot

Qualcomm used Investor Day to reframe itself as an AI infrastructure company. The targets are aggressive: a $40B non-handset business by fiscal 2029, more than $15B from data centers, and a software stack broad enough to challenge the idea that AI silicon has to be vertically owned by one vendor.

FY2029 target: $40BData center target: $15B+

Read article

AI memory supply chain, HBM, and data center bottleneck graphic

Semiconductors / Memory

Jun 24, 2026

MU•10 min read

Micron's Record Q3 Shows AI Memory Is Still the Tightest Bottleneck

Micron reported record fiscal Q3 2026 results, with revenue of $41.46 billion, non-GAAP EPS of $25.11, and a fiscal Q4 outlook for $50.0 billion of revenue. The read-through is that memory is no longer a commodity footnote in the AI cycle; it is one of the scarcest and most strategically valuable parts of the stack.

Revenue: $41.46BNon-GAAP EPS: $25.11

Read article

2026-06-22

Abstract memory supply chain graphic connecting Micron and Anthropic

AI Infrastructure

Jun 22, 2026

MU•9 min read

Micron Anthropic Pact

Micron's new strategic agreement with Anthropic turns memory and storage into a front-line AI infrastructure input. The deal ties HBM, DRAM, and SSD supply to one of the most capital-intensive frontier labs while Micron also takes an equity stake in Anthropic's latest funding round.

Memory scope: HBM / DRAM / SSDAnthropic valuation: $965B

NVIDIA Halos for Robotics Turns Physical AI Safety Into a Platform

NVIDIA says Halos for Robotics is the industry's first full-stack safety system for robotics and physical AI, built on 18,600+ engineering years of autonomous-vehicle safety work and already adopted first by Agility Robotics. The market implication is that the next leg of physical AI will be decided less by raw model skill than by who can package compute, sensors, inspection, and certification into one deployable platform.

Engineering base: 18,600+ yrsFirst adopter: Agility

Read article

2026-06-16

Intel foundry process node, risk production, and power-performance graphic

Semiconductors / Foundry

Jun 16, 2026

INTC•10 min read

Intel 18A-P Risk Production Makes Foundry Execution the Real Test

Intel says 18A-P is now in risk production with 9% better performance, 18% lower power, and improved thermal and via resistance versus 18A, while remaining design-rule compatible. The market read-through is simple: Intel is trying to turn foundry credibility into a real customer win, but the equity case still depends on whether it can execute faster than the AI demand curve moves.

Performance: +9%Power: -18%

Read article

2026-06-08

AI compute, sovereign infrastructure, and university research graphic

Semiconductors / AI

Jun 8, 2026

AMD•10 min read

AMD's £2 Billion UK AI Push Turns Compute Into Industrial Policy

AMD says it will invest up to £2 billion over five years in the United Kingdom, pairing sovereign AI infrastructure with university partnerships and photonic networking research. The market implication is bigger than one country: compute is becoming a geopolitical asset, and the companies that can sell the full stack get more leverage than the ones selling chips alone.

Investment: £2bnResearch partners: 2+

Read article

What to expect

Evidence-first notes with a visible point of view.

This section collects sharp takes on earnings, shareholder meetings, and market structure. Each new piece should make the thesis, the facts, and the implications obvious within the first few screens.

Expect direct analysis, not generic commentary.

Expect the data to be explicit and the argument to be easy to follow.

About Us•Privacy Policy•Terms of Service•Join our Slack