AMD Expands Taiwan CPU Capacity; AI Washing Wave Puts Investors on Alert

AMD expands Taiwan capacity as AI servers tighten CPU supply, while PR firms warn of a growing AI rebranding wave.

🗄️ AMD Ramps Up Taiwan Production as Global CPU Market Tightens

Decoded: Advanced Micro Devices is expanding production capacity at its Taiwan supply chain — where TSMC manufactures AMD's latest-generation EPYC server processors — as the global CPU market tightens, Reuters reported May 21. Demand pressure stems from the AI data center build-out: hyperscaler rack architectures require substantial CPU resources alongside GPU accelerators to manage inference pipelines, memory coordination, and inter-chip communication. AMD's EPYC server CPU line has gained data center market share for the third consecutive year, narrowing the gap with Intel's Xeon platform. TSMC's advanced packaging capacity — already under pressure from Nvidia's Blackwell GPU ramp — has made early production allocation a competitive priority for chip designers reliant on cutting-edge nodes. (Reuters, May 21, 2026)

Why it matters: AMD's capacity expansion is a read-through on CPU demand durability: if hyperscalers were pulling back on AI infrastructure spending, they would not be driving CPU tightness alongside GPU shortages. For investors, the CPU tightening dynamic also reshapes the Intel vs. AMD server battle — AMD is securing TSMC allocation at the same time Intel is attempting to recover process node competitiveness with Intel 18A. An AMD that locks in Taiwan capacity in a constrained market gains pricing power and delivery certainty that can translate into enterprise contract wins. The CPU layer of AI infrastructure is less visible than GPU procurement but is foundational to every large-scale inference deployment.

📊 PR Firms Flag Wave of AI Washing as Companies Rebrand to Claim AI Expertise

Decoded: UK public relations firms are warning that companies across sectors are performing what one industry source described as "yoga-level stretches" to rebrand themselves as AI specialists, capitalizing on investor appetite for AI-associated businesses, The Guardian reported May 24. The pattern echoes the dot-com rebranding wave of 1999–2000, when companies appended ".com" to their names to capture valuation premiums regardless of internet capability. PR professionals noted that companies with minimal AI deployment — including logistics, financial services, and HR software firms — are updating investor materials and press releases to emphasize AI capabilities they have not yet built. Regulators in the UK and EU are reportedly examining whether AI-capability claims in corporate communications constitute misleading investor disclosure. (The Guardian, May 24, 2026)

Why it matters: AI washing presents a material risk for investors building AI-themed portfolios. Companies claiming AI capabilities without supporting revenue, compute infrastructure, or engineering teams will face valuation compression when quarterly results fail to reflect AI-driven growth. For active investors, the bifurcation between companies with genuine AI revenue — Nvidia (NVDA), Palantir, Microsoft, Alphabet — and companies using AI language without AI revenue is widening. If regulatory scrutiny of AI capability disclosures follows the SEC's 2023 ESG enforcement pattern, auditable evidence for AI claims will become a compliance requirement, adding cost and reputational risk to firms that have overstated their positions.

Stay decoded. See you tomorrow.

— The Get AI Decoded Team