The AI Stock Nobody's Talking About That Could Be This Year's Biggest Winner

Qualcomm's fiscal Q2 revenue and adjusted earnings declined year over year, with smartphone revenue down 13% to $6 billion, but the stock rallied after management highlighted strong interest in custom AI ASICs and multiple customer wins in AI data center chips and inference accelerators. The company is targeting AI inference applications and edge AI as growth drivers beyond smartphones, while its automotive revenue rose 38% year over year to $1.33 billion. Shares trade at 19x earnings, suggesting upside if its AI chip pipeline converts into revenue.

Analysis

The market is beginning to re-rate Qualcomm not on handset cyclicality but on the probability that it becomes a viable second-source supplier in the AI inference stack. That matters because inference is a much broader TAM than frontier training: it expands the buyer universe from a handful of hyperscalers to enterprise, automotive, and edge deployments, where Qualcomm already has distribution and architectural credibility. If even one or two design wins scale into multi-generation sockets, the stock’s current multiple likely understates optionality rather than just cheapness.

The second-order winner is the broader edge/embedded ecosystem: Qualcomm’s push reinforces a split architecture where training stays concentrated and inference gets pushed closer to the user/device. That is constructive for OEMs and system integrators that need lower-power silicon, but it is a competitive threat to parts of Nvidia’s moat if inference workloads migrate to custom ASICs optimized for cost per token rather than raw performance. It also raises pressure on Intel’s foundry ambitions and on merchant accelerator vendors that rely on generalized enterprise demand.

The key risk is that the current enthusiasm prices in narrative before it prices in silicon economics. Custom AI chips have long design-to-revenue lags, and a few “wins” can remain immaterial for 12-18 months if deployment volumes or margins disappoint. The setup could reverse quickly if management’s Investor Day details are vague, if hyperscaler interest proves non-exclusive, or if handset weakness worsens enough to offset early AI contributions.

Contrarian view: the market may be underestimating how much of the near-term upside is in the sentiment delta, not the earnings delta. If Qualcomm can validate even a narrow pipeline of inference ASICs, the multiple can expand before revenue inflects; if it cannot, the stock is still vulnerable because smartphone earnings remain the core bridge. This makes the trade more about catalyst sequencing than fundamental conviction today.

Key Decisions for Investors

• Go long QCOM into Investor Day on a 1-3 month horizon, but size for event risk: the stock can re-rate 10-20% on credible AI pipeline disclosure even before revenue shows up.
• Use a call spread in QCOM rather than outright equity to express upside optionality with capped premium risk; best if implied vol remains elevated but not extreme ahead of the catalyst.
• Pair trade: long QCOM / short a basket of handset-exposed semiconductor peers over 3-6 months, betting that AI optionality gets a relative multiple premium while pure smartphone names remain structurally challenged.
• If the company provides only vague customer language, fade the move via a tight stop or short-dated put spread; the downside is sharper than consensus expects because the core earnings engine is still handset-driven.
• Monitor NVDA on any sign that inference optimization shifts spend toward custom silicon; if that theme gains traction, consider trimming high-multiple AI beneficiaries and rotating toward edge/ASIC enablers.