Move Over Marvell, Here Are the Next 2 $1 Trillion Semiconductor Stocks

The article argues that AMD and ASML have a much clearer path to a $1 trillion market cap than Marvell, citing AMD's $865 billion valuation and ASML's roughly $710 billion valuation. AMD is benefiting from inference and agentic AI demand, plus its MEXT acquisition and data center CPU growth, while ASML remains essential to AI-chip and HBM manufacturing through its EUV monopoly. Marvell is also growing quickly, with revenue expected to rise 40% to $11.5 billion, but its sub-$250 billion market cap and nearly 70x forward P/E make the trillion-dollar case more speculative.

Analysis

The market is rewarding the picks-and-shovels of AI more than the chip designers, but the second-order implication is that the bottleneck is shifting from compute to memory density and interconnect bandwidth. That makes ASML the cleanest structural winner: if HBM and advanced-node capacity keep expanding, lithography demand compounds across both logic and memory, and its pricing power is less exposed to customer-design wins/losses than fabless names. The key takeaway is that the AI capex stack is becoming more modular, which should compress the moat of integrated platform narratives like NVIDIA’s ecosystem over time while widening the moat of equipment and memory-optimization vendors.

AMD’s path to the trillion mark is the most plausible near-term because it is already close enough that execution only needs to stay above a high bar, not create a new category. But the market may be underestimating how much of the upside is already priced into the AI inference/agentic AI narrative; if enterprise rollouts slip by 1-2 quarters, the multiple can de-rate faster than revenue ramps. The higher-conviction angle is not absolute share gains versus NVIDIA so much as a mix shift toward CPU-heavy agent workloads, where AMD’s product architecture and memory-management acquisition can improve attach rates in existing data center refresh cycles.

Marvell is the cleanest “good business, too early” setup. Its custom silicon and optical connectivity are strategically important, but the downside case is that hyperscalers keep dual-sourcing or internalizing more design content, which caps long-run margin leverage even if revenue stays strong. In other words, the company can be a beneficiary of AI capex without being the primary equity winner; that usually translates into episodic upside on beats, but lower terminal multiple than the market wants to pay.

Key Decisions for Investors

• Long ASML on a 6-12 month horizon; best risk/reward in the group because it monetizes both AI logic and HBM buildout without needing end-demand share gains. Use call spreads rather than outright equity to express upside into a possible re-rating.
• Hold/accumulate AMD only on pullbacks or after confirmation of inference deployment traction; near-term upside is capped if the market already discounts the trillion-dollar narrative. Prefer a call spread funded by selling higher strikes to reduce multiple-compression risk.
• Fade MRVL strength into rallies with a tactical short or covered-call overlay over 1-3 months; the business is strategically relevant, but customer concentration and possible design displacement argue for lower terminal multiple than consensus implies.
• Pair trade long ASML / short MRVL for 3-6 months as a cleaner expression of AI infrastructure scarcity versus AI-adjacent execution risk.
• For traders, buy NVDA downside-protected exposure only if ecosystem strength translates into visible custom silicon monetization; otherwise the marginal upside may accrue to suppliers and not the platform leader.