AI Spending Could Soar 500%: 3 Brilliant AI Stocks to Buy (Hint: Not Nvidia)

The article identifies Broadcom, AMD, and Taiwan Semiconductor Manufacturing (TSMC) as key beneficiaries of the projected fivefold increase in AI infrastructure spending. Broadcom is capitalizing on the demand for custom AI chips (ASICs) from hyperscalers, securing major design wins with Alphabet, Meta, ByteDance, and a recent $10 billion order from a fourth customer, projecting a $60-90 billion opportunity by FY2027 from its initial three clients. AMD is positioned for growth in AI inference with its GPUs and ROCm software, gaining adoption among major AI players and contributing to open interconnect standards. TSMC is highlighted as a foundational winner, benefiting universally from AI chip demand due to its unparalleled leadership in advanced semiconductor manufacturing and strong pricing power, with AI chip demand expected to grow over 40% CAGR through 2028.

Analysis

The projected fivefold expansion in AI infrastructure spending is creating significant growth vectors for semiconductor firms beyond the current market leader, Nvidia. Broadcom (AVGO) is capitalizing on the demand for custom ASICs from hyperscalers seeking to reduce costs and diversify supply chains. The company's established success with Alphabet's TPUs has led to major design wins with Meta and ByteDance, creating a management-projected $60 to $90 billion revenue opportunity by fiscal 2027. This momentum is accelerated by a new, surprisingly fast-tracked $10 billion order from a fourth major customer for delivery next year. Concurrently, Advanced Micro Devices (AMD) is carving out a niche in the AI inference market with its GPUs, presenting a viable alternative for companies not pursuing full custom silicon. AMD's traction is evidenced by its hardware being used by seven of the top ten AI players and the development of its ROCm 7 software platform, which is considered sufficient for inference workloads. The UALink Consortium, which includes AMD, further aims to challenge Nvidia's closed ecosystem by promoting an open interconnect standard. Foundational to this entire ecosystem expansion is Taiwan Semiconductor Manufacturing (TSM), which operates as a de-facto winner regardless of which chip designer gains market share. TSM's manufacturing leadership in advanced nodes, where competitors have struggled, grants it significant pricing power, highlighted by reports of a planned 10% price increase next year. TSM management reinforces this outlook, forecasting AI chip demand to grow at a compounded annual rate exceeding 40% through 2028.

Key Decisions for Investors

• Investors should monitor Broadcom's custom silicon revenue and announcements of new hyperscaler partnerships as key indicators of its ability to execute on its multi-billion dollar AI pipeline.
• Consider AMD as a higher-risk, higher-reward play on AI market diversification, with the adoption rate of its ROCm software and progress within the UALink Consortium serving as critical catalysts for market share gains against Nvidia.
• View TSMC as a foundational, lower-risk holding for broad exposure to the AI infrastructure build-out, as its manufacturing dominance and pricing power position it to benefit universally from rising chip demand.
• Given the competitive landscape, a balanced strategy could involve using TSM for core AI exposure while selectively adding positions in AVGO or AMD to target specific growth opportunities in custom chips and the merchant GPU market, respectively.