Prediction: A $5,000 Investment Split Between Nvidia and Broadcom Will Triple by 2028

The article argues that Broadcom and Nvidia could roughly triple or more in value by 2028, based on AI growth assumptions rather than a reported new event. Broadcom CEO Hock Tan expects AI semiconductor revenue of $115 billion next year and $230 billion in 2028; the author’s scenario adds $65 billion in non-AI semiconductor revenue and assumes a 50% profit margin and 35× earnings valuation to imply a $5.16 trillion market cap, versus about $1.7 trillion today. Nvidia management expects 70% revenue growth for its next fiscal year; the author extrapolates 70% growth for two years to $1.2 trillion in revenue and a market cap above $20 trillion, assuming a 55% margin and 35× earnings.

Analysis

The article’s valuation case is more fragile than its demand case: it extends management growth expectations into future periods, then assumes exceptionally high net margins and an unchanged premium earnings multiple. Those are independent assumptions, not a single forecast. A modest miss in any one can erase much of the implied upside; verify whether the cited AI revenue targets are backed by customer commitments, capacity reservations, and deliverable supply—not just pipeline visibility.

The second-order risk differs by company. Broadcom’s custom-chip opportunity can expand AI compute economics, but it may also concentrate revenue among a small number of hyperscalers with bargaining power, while foundry, advanced-packaging, and networking capacity constrain shipment conversion. Custom silicon could shift some workloads from Nvidia, but the addressable mix depends on workload-specific performance and software costs; it is not automatically a one-for-one GPU replacement. Nvidia’s broader platform may benefit from workload flexibility, yet hyperscaler returns on AI capex—not chip demand alone—ultimately determine whether spending persists.

Near term, sentiment and positioning can drive both stocks more than long-range arithmetic. Over 1–3 months, watch company guidance, customer capex commentary, and evidence of shipment conversion. Over 6–18 months, monitor AI infrastructure utilization, customer concentration, supply constraints, and gross/net margin trajectory. The thesis weakens if capex plans are deferred, AI revenue growth decelerates materially, or margins fall while capacity ramps.

Key Decisions for Investors

• Do not underwrite a threefold return from the article’s spreadsheet. Before adding exposure, reconcile the cited revenue targets with company filings and the relevant fiscal periods, and establish whether “profit margin” means net margin; the valuation argument depends heavily on that distinction.
• For a directional AI trade, prefer staged exposure to NVDA over an equal-weight basket at current sentiment levels; add only after earnings confirm shipment growth and customer demand. This expresses demand breadth while limiting reliance on a few custom-chip programs. Falsify if guidance or order commentary shows a material growth slowdown.
• Keep AVGO as a smaller, catalyst-driven position rather than treating its long-range target as bankable. Track customer concentration, supply commitments, and AI semiconductor revenue conversion; reduce exposure if delivery timing slips or margin performance contradicts the earnings assumptions.
• Avoid chasing both names immediately after bullish coverage. Reassess after the next earnings and hyperscaler capex updates; if spending is intact but either stock sells off on results, distinguish valuation compression from deteriorating orders before adding.