Tuesday was ugly across the board.

Qualcomm dropped 8% on June 9th, touching a session low of $192.67 before clawing back to close near $205. Volume hit roughly 29.9 million shares against a 22-million average. Marvell got taken out back and shot down 10%. Broadcom caught stray fire too. The whole semiconductor sector got treated like one trade, one thesis, one risk. That is where the market made a mistake worth paying attention to.

The trigger was ByteDance. News circulated that the company is moving ahead with a custom AI silicon agreement that touches Qualcomm directly. Here is the thing though: the stock had already surged roughly 60% over the prior month after that same ByteDance ASIC deal was first reported. So some of Tuesday’s drop was pure profit-taking dressed up as fear. The rest was contagion. Markets saw “AI chip” and “China” in the same sentence and hit sell first, asked questions later.

What gets lost in that reaction is what Qualcomm actually builds.

This is not an Nvidia story. It is not a data center GPU story. Qualcomm’s core business is Snapdragon processors, edge AI silicon, automotive compute, IoT platforms, and mobile connectivity. The U.S. export restrictions that have been tightening around AI hardware are specifically designed around high-bandwidth server accelerators that exceed certain performance thresholds. Qualcomm’s chip portfolio does not live in that category. Its competitive advantage sits at the edge of the network, in the phone in your pocket, the car on the highway, the industrial robot on the factory floor. Completely different products. Completely different regulatory exposure. The market sold the sector as a monolith, which is a blunt instrument and usually a temporary one.

Slight tangent, but it actually matters here: the global edge AI hardware market is estimated at around $26 billion in 2025 and is projected to hit $58.9 billion by 2030, compounding at roughly 17.6% annually. That is not a speculative number. It is showing up in Qualcomm’s actual reported line items right now, not just in analyst models.

The numbers tell that story directly.

• FY2025 revenue: $44.28 billion, up 13.66% year over year
• Q4 FY2025 revenue: $11.3 billion, up 10% YoY, beating consensus by roughly 4.6%
• Automotive and IoT combined grew 27% in FY2025
• Automotive alone hit $1.1 billion in Q1 FY2026, up 15% YoY
• IoT revenue: $6.6 billion for full FY2025, up 22% YoY
• Free cash flow: approximately $10 billion annually
• Forward P/E: roughly 21.9x at current levels
• 52-week range: $121.99 to $259.92
• Dividend yield: approximately 1.7%
• JPMorgan raised its price target to $265 ahead of the June 24 Investor Day

CEO Cristiano Amon has publicly stated a target of $22 billion in non-handset revenues by 2029. Automotive and IoT are already growing double digits. AI PCs running Snapdragon X are gaining share. The Dragonwing platform for robotics is a brand-new revenue layer that barely exists in the current financials yet.

Now the honest part.

China exposure is the real risk and it deserves straight talk. Somewhere between 46% and 65% of Qualcomm’s revenue connects back to the Chinese ecosystem depending on how you account for it. Any escalation in trade tensions, any acceleration of domestic substitution through HiSilicon or similar, and that number becomes a serious problem. Apple’s gradual shift away from Qualcomm modems is a slow overhang that has not gone away. And Nvidia’s RTX Spark chip targeting Windows on Arm PCs is real competition entering the AI PC segment. None of these risks are new. The stock has been living with them for months. But they are worth naming clearly before anyone adds exposure here.

What matters is the June 24 Investor Day and what Amon puts on the table.

Management is expected to detail data center inference opportunities, custom silicon shipments to at least one hyperscaler beginning late 2026, and a fuller picture of the automotive and robotics pipeline. If those details land with specificity and credibility, the street will have a harder time treating QCOM like a generic AI hardware trade. If the event disappoints or feels like slides without substance, the $192 to $200 range becomes a much more important conversation. Start small if at all. Let June 24 do the work of confirming or killing the thesis before sizing up.

The market sold Qualcomm because it was nearby when fear hit the sector. The underlying business grew 13.66% last year. Automotive and IoT are compounding. Free cash flow is running near $10 billion annually. These facts did not change on Tuesday. What changed was sentiment, and sentiment at this speed tends to overshoot in both directions.

Whether $205 holds or we revisit the low end of that range before June 24 is genuinely unclear. That uncertainty is not a reason to ignore it.

Full breakdown here if you want to go deeper before the investor day.

– The Cheap Investor