Is Broadcom in Trouble Now That Alphabet Is Getting Chips From Marvell Too?

Broadcom shares have fallen >10% in under two weeks after Alphabet expanded its long-term AI chip design partnership with Marvell, raising concerns Broadcom could be pressured on pricing or share. The article argues the dip is overdone given fundamentals: Broadcom’s AI semiconductor segment revenue rose 143% year over year, and management guided for AI semiconductor revenue to more than triple YoY in fiscal Q3. It also notes overall revenue is expected at $29.4B in fiscal Q3 (+84% YoY) while emphasizing Alphabet’s likely continued use of multiple suppliers.

Analysis

Alphabet’s added supplier relationship is better read as procurement optionality than displacement. In hyperscaler silicon, more vendors usually means better pricing discipline and lower dependency risk, not a sudden loss of end-demand; the economic prize is still expanding AI capex, which tends to lift the whole custom-chip stack before it hurts the leader.

For Broadcom, the market reaction looks more like multiple compression on “share loss” fears than a revision to revenue math. The real risk is not volume collapse but incremental ASP pressure and a slower slope of margin expansion if customers become more effective negotiators. Marvell gets narrative upside, but it still needs proof that these design wins convert into meaningful revenue before the next cycle; otherwise it remains a smaller beneficiary of a larger market.

The key catalyst window is the next 1-3 months, when hyperscaler capex commentary and Broadcom’s AI revenue guide will either validate or puncture the selloff. Falsifiers are simple: if Broadcom’s AI cadence holds and Alphabet’s spending continues to rise, the dip is buyable; if Alphabet explicitly starts substituting away from custom silicon or Broadcom trims its AI outlook, the bear case becomes real. The contrarian miss is that diversification is being confused with substitution—those are not the same thing in a supply-constrained AI buildout.

Key Decisions for Investors

• Buy AVGO on further weakness over the next 1-2 sessions; use the selloff as a sentiment event, not a fundamentals event. Risk/reward favors a rebound if the market does not get a confirming negative guide from hyperscaler capex.
• Initiate a relative-value pair: long AVGO / short MRVL into any Marvell-led bounce. The thesis is that optionality premium gets assigned to the smaller name, while the incumbent with larger AI revenue base and better margin leverage should outperform on a 1-3 month horizon.
• Do not chase MRVL strength solely on this partnership headline; wait for a disclosed revenue ramp or margin impact. Absent that, the move is mostly narrative and vulnerable to reversal once the market asks for bookings and conversion timing.
• Set an alert for the next GOOG/GOOGL capex update and AVGO AI revenue guide. If both remain constructive, add to AVGO; if either signals slower AI spend or lower custom-silicon intensity, cut exposure quickly.