Broadcom's AI chip revenue surges 106% year-over-year to $8.4 billion, outpacing Microsoft's slower cloud growth despite higher spending

Broadcom reported fiscal Q1 2026 (ended Feb. 1, 2026) AI semiconductor revenue of $8.4 billion, up 106% year-over-year, with management guiding Q2 AI revenue to reach $10.7 billion (140% growth). Microsoft's intelligent cloud segment (which includes Azure) grew 29% year-over-year to $32.9 billion in fiscal Q2 2026, but the company spent $37.5 billion on capital expenditures to build data centers.

Broadcom's momentum comes from deep, multi-year custom silicon projects with major cloud providers (what the CEO calls 'deep, strategic, and multi-year' collaborations), giving the company locked-in visibility into customer demand through 2028. Microsoft faces the opposite pressure: its CFO acknowledged that demand still outpaces available infrastructure despite aggressive spending, meaning the company must continuously invest just to keep up.

For investors choosing between the two: Broadcom trades at a forward price-to-earnings ratio of 38 (much lower than its trailing ratio of 82, showing how fast earnings are growing), while Microsoft trades at a trailing ratio of 27. An analyst argues Broadcom's faster growth and customer lock-in justify the premium, making it the better bet on AI infrastructure spending — especially for investors who want exposure to the chips and networking gear powering AI, rather than the software and services running on top of it.

The risk to watch: Broadcom depends heavily on a handful of hyperscalers (the largest cloud companies); losing even one major customer could sharply reverse its trajectory. Microsoft's risks are different — potential disruption from AI replacing software jobs, and the ongoing cost of building out cloud capacity.