Prediction: This Is What a $5,000 Investment in Broadcom Will Be Worth by 2028

Broadcom forecasts AI semiconductor revenue of $230 billion in fiscal 2028, up from projected $65 billion in FY2026 and $115 billion in FY2027, driven by custom AI-chip demand from hyperscalers including Alphabet, Meta, OpenAI, and Anthropic. The company reported AI semiconductor revenue of $16.7 billion, up 221% year over year, and guided for $21.7 billion next quarter, up 236%. The article estimates that reaching the FY2028 target could lift total revenue to roughly $280 billion and, assuming a 45% margin and 30x earnings multiple, support a $3.78 trillion valuation versus the current $2.7 trillion market capitalization.

Analysis

The investable question is not whether custom silicon grows, but whether AVGO can sustain merchant-like economics while becoming increasingly dependent on a small number of internally funded hyperscaler programs. ASICs lower customers’ cost per token only after workload maturity and sufficient utilization; that makes adoption strongest in inference and repeatable internal workloads, while NVDA retains a defensible position in frontier training, rapidly changing models, and software portability. The second-order beneficiaries are TSM, MU and SK Hynix through leading-edge wafer, advanced-packaging and HBM content, although those supply constraints can also cap AVGO shipment upside over the next 12-18 months.

The market should discount the long-range revenue target more heavily than the bullish narrative implies: it requires sustained hyperscaler capex growth, successful tape-outs, adequate HBM/CoWoS capacity, and no material customer insourcing beyond AVGO’s design role. Custom-chip programs also create lumpy revenue recognition and unusually high customer-concentration risk; a single delayed deployment can produce a sharp guide reset even if end demand remains intact. Near-term upside depends on order visibility and supply allocations, but over 6-18 months the key debate shifts to mix-adjusted gross margin and whether incremental AI revenue converts to FCF at a rate consistent with the valuation premium.

Key Decisions for Investors

• Maintain a tactical long AVGO only on post-earnings confirmation that next-12-month AI backlog and advanced-packaging supply allocations are rising; use a 10-12% downside stop from entry and target 20-25% upside over 6-12 months. Do not underwrite the position to the out-year revenue aspiration absent disclosed customer commitments.
• Express the custom-silicon theme through a basket long TSM and MU versus short SOXX over 3-6 months, sized modestly: TSM/MU capture foundry, packaging and HBM content while SOXX carries broader exposure to cyclical analog and legacy semiconductor demand. Exit if HBM pricing weakens materially or leading-edge foundry utilization guidance falls.
• Avoid a directional short NVDA solely on ASIC adoption. A cleaner relative-value watch is long AVGO / short AMD after evidence that hyperscaler inference deployments are displacing merchant accelerators; require AMD data-center GPU guidance or major customer commentary as confirmation before initiating.
• Set an earnings watch on AVGO for AI segment gross margin, top-customer concentration, cancellation provisions and cash-conversion trends. A guide cut, weaker AI margin despite revenue growth, or delayed supply qualification would falsify the operating-leverage thesis and warrant reducing exposure immediately.