Which Growth Stock Is a Better Buy: Nvidia or Broadcom?

Nvidia posted fiscal Q1 revenue of $81.6 billion, up 85% year over year, while Broadcom’s AI semiconductor revenue surged 143% to $10.8 billion and management guided to about $16 billion in AI chip revenue this quarter. Broadcom’s stock fell despite record results after management merely reiterated its fiscal 2027 AI revenue target of more than $100 billion, but the article argues the post-earnings selloff makes the stock more attractive versus Nvidia’s ~25x forward P/E and Broadcom’s ~27x. The piece is bullish on AI infrastructure demand overall, with Broadcom seen as better positioned on the custom-silicon trend.

Analysis

The key second-order shift is not simply that AI demand is strong, but that hyperscalers are monetizing supply-chain leverage by dual-sourcing and vertically integrating. That creates a richer market for custom silicon and networking than for merchant GPUs, but it also means the value pool is fragmenting: the near-term beneficiary is still the incumbent with the best software moat, while the medium-term winners are the firms that sit inside the customer’s architecture and recapture design wins after each new generation.

For NVDA, the risk is less an abrupt demand cliff than a gradual erosion of pricing power as alternatives become credible over the next 4-8 quarters. If capex keeps growing but average selling prices normalize, revenue can stay resilient while margin multiple compression does the damage. AVGO faces the mirror image: its AI mix can compound faster, but customer concentration means one delayed tape-out or a decision to multi-source can interrupt the growth slope and compress sentiment quickly, even if the fundamental story remains intact.

The market reaction looks more interesting than the headline guidance because it suggests investors are rewarding acceleration more than absolute scale. That tends to favor pullback buyers in the better-placed structural compounder, but only if they accept that the stock is already discounting a long runway. The contrarian miss is that “AI infrastructure” is no longer a single trade; it is a supplier-power trade, and the next leg likely comes from whoever owns the bottleneck between model training, network throughput, and custom chip integration rather than the largest GPU supplier alone.

Key Decisions for Investors

• Go long AVGO on further post-earnings weakness over the next 1-3 weeks; use the drop to build a position for a 6-12 month hold. Risk/reward favors a re-rating if management delivers even a modest upward revision next quarter; downside is limited unless a top customer signals a design loss.
• Reduce/hedge NVDA into strength over the next 1-2 months by buying put spreads or trimming exposure. The trade is not against demand, but against multiple compression if hyperscaler custom silicon adoption continues to broaden.
• Pair trade: long AVGO / short NVDA for 3-6 months. This expresses the thesis that custom accelerators and networking are taking share of wallet from merchant GPUs while still staying long the AI capex cycle; invalidate if NVDA re-accelerates pricing power or AVGO gross margins start to deteriorate faster than expected.
• For lower-cost convexity, buy AVGO call spreads on a 3-6 month horizon rather than outright shares. The setup benefits from a sentiment reset after the selloff, with defined risk if AI revenue guidance stalls again.
• Monitor GOOGL and META as indirect beneficiaries over the next 2-4 quarters; they are better positioned than most to internalize compute economics through in-house silicon. Consider adding on weakness if the market overreacts to any near-term AI capex concerns.