Lisa Su's 10-Year AI Supercycle: 3 Key Stocks for the Future of Compute

Lisa Su's AI Supercycle Vision

• 💡 AMD CEO Dr. Lisa Su predicts a 10-year AI supercycle, stating we are only two years into it, rejecting the idea of an AI bubble.
• 📈 This supercycle is driven by actual compute demand, revenue, profitability, and global data center expansion, not speculative valuations.
• ⚠️ The current AI investment wave is fueled by shortages of GPUs, advanced packaging, HBM memory, and power infrastructure.
• 🚀 AI workloads are doubling every year, accelerating faster than any previous technology adoption curve, including smartphones and cloud computing.

AMD: Powering AI Compute

• ✅ Advanced Micro Devices (AMD) is executing the AI supercycle vision, transforming into a data center and AI acceleration leader.
• 📊 AMD reported record quarterly revenue of $9.2 billion, with its data center business growing above 50% and on track to exceed 60%.
• 🧠 The Instinct MI300 GPUs are AMD's fastest-ramping product, integrated by major hyperscalers, and AMD partners with OpenAI for compute.
• 🛠️ AMD's competitive edge includes chiplet architecture and expansion into full-stack infrastructure, aiming for double-digit share of the $1 trillion data center market by 2030.

Micron: Essential AI Memory

• 💾 Micron Technology (MU) is critical to the AI buildout because AI is memory-bound, requiring huge capacity, speed, and bandwidth.
• 🎯 Micron is redirecting capacity towards data center AI memory, exiting lower-margin consumer memory by early 2026.
• 📈 HBM revenue hit $2 billion last quarter, and Micron holds 21% global HBM market share, surpassing Samsung.
• 🇺🇸 As the only US-based HBM supplier, Micron is strategically vital for American AI sovereignty and national security initiatives.

Marvell: High-Speed AI Networking

• 🌐 Marvell Technology (MRVL) is a crucial, lesser-known hero, addressing the networking bottleneck in scaling AI clusters.
• 🚀 Marvell's data center revenue jumped from $2.5 billion to over $5.8 billion, with management forecasting 45% AI-driven growth in fiscal year 2026.
• 🤝 The company acquired Celestial AI for $3.25 billion, adding photonic fabric technology for multi-rack optical connections in generative AI infrastructure.
• 🛣️ Marvell is building the roads and highways inside tomorrow's AI data centers, becoming indispensable infrastructure.

Pillars of AI Infrastructure

• 🏗️ AMD, Micron, and Marvell are identified as the pillars of the AI energy grid, building the brains, supplying memory, and enabling connectivity.
• 🌍 This represents the early phase of a historic infrastructure transformation, where data centers evolve into AI factories.
• 💰 Long-term investors should focus on structural demand backed by irrefutable math in this decade of compute.