Qualcomm Is Remaking Itself Into an AI Company. Its Shares Look Dirt Cheap.

Qualcomm recently bought AI software company Modular in an all-stock deal valued at about $4 billion and also secured a multi-year agreement with Meta for data center processors. Management said AI infrastructure revenue could exceed $15 billion by fiscal 2029, up from essentially nothing today, while the stock still trades at a P/E of 21 versus a 44 tech-sector average. The article is broadly bullish on Qualcomm’s AI pivot, though execution risk remains.

Analysis

QCOM is becoming a higher-beta AI optionality vehicle, but the market is still valuing it like a mature handset supplier. That disconnect matters because the stock does not need to become a dominant AI platform to rerate; it only needs credible evidence that data-center inference is becoming a meaningful second engine over the next 12-24 months. The Modular acquisition is strategically important not for revenue on day one, but because it lowers the software friction that usually keeps custom silicon from scaling beyond a few lighthouse customers.

The competitive implication is less about beating NVDA head-on and more about attacking the portion of spend where GPU economics are weakest: inference, latency-sensitive workloads, and edge-to-cloud orchestration. If QCOM can pair a software layer with decent hardware economics, the real second-order winner is hyperscale buyers, who gain leverage in procurement and are likely to push more custom silicon bids across the industry. That creates mild pressure on NVDA’s pricing power at the margin, while increasing the probability that other accelerator vendors and CPU incumbents are forced into more aggressive bundling.

The key risk is execution timing. The revenue narrative is likely too early for near-term fundamentals, so the stock may have already discounted the strategic shift before the P&L catches up. A miss on design-win cadence, delayed integration of the software asset, or Meta remaining a single-customer proof point would likely compress the multiple back toward core-semiconductor levels rather than keep expanding it.

The contrarian view is that consensus may be underestimating how cheap QCOM is relative to its option value. At current valuation, the market seems to be pricing the AI strategy as a low-probability call option instead of a base-case growth reacceleration, which creates attractive asymmetry if management can show even modest traction over the next two earnings cycles.

Key Decisions for Investors

• Long QCOM on a 6-12 month horizon; use pullbacks to build a position, targeting a rerating if the market starts capitalizing AI infrastructure revenue at even a fraction of software-like multiples. Risk: if design wins stay anecdotal, the stock likely reverts to low-20s P/E.
• Pair trade: long QCOM / short NVDA as a relative-value hedge over 3-6 months. Thesis: QCOM offers earlier multiple expansion from a lower starting point, while NVDA faces more pressure if hyperscalers keep shifting a portion of inference workloads to custom silicon. Risk: NVDA can still outperform on capex momentum and model training demand.
• Buy META on weakness only if management commentary confirms QCOM is part of a broader custom-silicon strategy, not a one-off procurement test. Time horizon: next 1-2 quarters. Risk/reward improves if the deal catalyzes broader vendor diversification rather than just optics.
• Avoid chasing QCOM after sharp post-announcement spikes; prefer selling cash-secured puts or call spreads to express bullishness with defined downside. This is a better structure while the market waits for concrete revenue conversion.
• Watch for a second design win outside Meta within the next two earnings reports; if absent, trim or hedge QCOM because the narrative may plateau before fundamentals catch up.