尽管短期面临手机业务逆风且估值偏高，QCOM 向汽车与 AI 数据中心的多元化布局仍将驱动长期增长。

On this week's edition, we're going to be revisiting Qualcomm. Now, Qualcomm is a global semiconductor, wireless technology, and increasingly becoming an AI computing infrastructure company.

Historically, Qualcomm has best been known for supplying Snapdragon processors, cellular modems to companies like Apple, radio frequency technologies, and wireless connectivity for company's actively trying to transform itself from a handset centered semiconductor company into a broader computing platform, spanning all kinds of devices, including smartphones and personal AI devices, automotive, industrial and consumer IoT devices, robotics, PCs, edge artificial intelligence, and now AI data centers as well.

Now, Qualcomm operates in two main segments. It's CDMA technology segment, also known as QCT, as well as its technology licensing division, known as QTL.

Now, under the QCT umbrella, this is where the company develops and sells semiconductor platforms. Its largest businesses are still handsets, automotive, and IoT devices. But, Qualcomm is now building out a significant data center product portfolio.

Now, under the QTL division, this is where the company's technology licensing monetization comes in. It has an enormous patent portfolio, including intellectual property essential to 5G wireless standards, and QTL also remains particularly important because licensing economics produce exceptionally high margins.

First off, Qualcomm faces competition from major companies in the semiconductor industry space. Now, when it comes to unique value, Qualcomm's core technological advantage remains high performance and low power consumption.

Now, this capability has developed through decades of designing chips like the Snapdragon processors for battery-powered smartphones where compute performance, artificial intelligence, connectivity, thermal management, and energy consumption all have to be optimized simultaneously.

Qualcomm increasingly believes that these same innovations have become extremely valuable for AI inference. Qualcomm wants to bring the architecture that made Snapdragon successful mobile devices into AI PCs, vehicles, robotics, industrial equipment, and eventually large-scale AI data centers.

Fiscal Q3 revenue came in around 9.95 billion versus 10.37 billion a year earlier. This represented a 4.4% decline year-over-year. But nonetheless, revenue did come in ahead of estimates, which were expected here around 9.7 billion dollars.

Now, also importantly, if you look at earnings, non-GAAP earnings were $2.21, which is actually representative of a 20% decline compared to last year, but nonetheless, only missed the estimates by about 2 cents, which was $2.23.

we can see that the handset revenue in its QCT division declined dramatically. They posted in this division $5.09 billion, down 20% year-over-year from $6.33 billion last year.

Now, management attributed much of this decline to lower chipset shipments as several major OEMs adjusted production plans because of memory supply constraints and higher component pricing.

This is extremely important because handsets still represent approximately 60% of the QCT division revenue for the quarter.

On the other hand, automotive has become one of Qualcomm's strongest businesses. Fiscal Q3 automotive revenue reached 1.59 billion, up an extraordinary 61% year-over-year from 984 million the previous year.

This also marked Qualcomm's 23 consecutive quarterly double-digit automotive revenue growth.

Management also acknowledged broad increases in semiconductor input costs, including wafer fabrication, advanced packaging, assembly, testing, as well as memory. Now, in response to that, Qualcomm did denote that they are raising prices, but those increases take time to flow through the financial statements.

Now, even recently this week on September 8th, Qualcomm announced a multi-generational collaboration with Amazon focused on AI data center infrastructure. The two companies will work together on customized silicon for AI inference and high-performance optical connectivity for Amazon's data centers.

Now, for Qualcomm, this provides arguably the strongest validation yet that its data center push is moving into real hyperscale deployments. Now, despite, like I said, the handset challenges, automotive revenue surged 61% year-over-year.

Now, that's meaningful because they're competing aggressively with Nvidia.

Qualcomm's diversification strategy is materializing with aggressive expansion in the data centers, automotive, and AI segments expected to drive non-handset revenue to $40 and earnings above $80 a share by fiscal year 2029.

Also, Qualcomm's low-power architecture could become particularly valuable as hyperscalers focus on increasingly energy costs and tokens per watt. And lastly, some other recent successes include multi-year partnerships and significant design wins in automotive.

For example, partnerships with BMW and Stellantis, as well as developments and partnerships with other data center deals, including Meta and Microsoft.

Now, these new deals also project strong non-handset growth and reinforces the company's AI leadership aspirations. First off, Qualcomm is losing Apple's modem business over time.

This remains the single biggest identifiable earnings headwind.

The Apple transition is occurring faster than originally anticipated, and handset revenue already declined 20% during Q3. Qualcomm must prove that its emerging businesses can replace that lost revenue.

Now, data center growth still carries execution risk. Amazon deal is significant, yes, but Qualcomm is entering one of the most competitive markets in technology.

Now, despite the exciting long-term AI story for Qualcomm, consensus fiscal 2026 revenue is still down approximately 3% to roughly $42.9 billion estimated, and the current consensus for fiscal year 2027 sees only $44.82 billion, only about 4.4% growth.

So, this means the present financial results have not yet caught up with the long-term AI narrative. And lastly, from an earnings multiple and valuation perspective, the stock is trading above its historical 5-year earnings multiple.

The current PE approximately on a forward basis is around 16.7 times, while its average is around 14.32 times.

Now, this is important while earnings and sales growth continue to track lower than their 5-year averages. All right, this is part of the segment where we look at the technical picture for Qualcomm.

Now, first off, Qualcomm has improved dramatically from the spring lows, but relative performance remains weak compared to the broader semiconductor group.

Now, Qualcomm's year-to-date return is roughly 4% gain versus approximately 12% for the S&P 500. Also, its 1-year return is approximately 8% positive versus approximately 8% for the S&P 500.

Now, in the near term, one could say that there's a bullish trend indicated with the price trading above its 20-day moving average, as well as the 50-day moving average. Now, we have to keep in mind that the 50-day moving average is still sloping down, which is something we have to be concerned about.

Yet, the stock is also trading near a next level of intermediate resistance level as indicated by the 100-period moving average. Now, from a momentum standpoint, the MACD histogram is now clearly positive and has been expanded.

This suggests upside momentum is strengthening rather than merely stabilizing.

Now, from a daily RSI indicator, this number has risen to approximately 67, which is a bullish indication, but because RSI remains below 70, the stock is not yet technically overbought by the traditional definition.

Therefore, this may give the current rally some additional room before momentum reaches an extreme level. Now, to summarize, until recently, investors could reasonably dismiss Qualcomm's data center ambitions as a distant diversification strategy.

The recent Amazon agreement makes that narrative a little bit more difficult.

AWS provides a major reference customer. Purchase commitments already exist, and Qualcomm expects revenues to begin in the December quarter. Now, Apple's decision to build its own modems chips to displace Qualcomm may be a medium to long-term handset headwind, but not a calamitous one, especially as Qualcomm is poised to grow in the automotive and Internet of Things.

Qualcomm has enduring strength in its silicon products and IP portfolio, and the company is poised to be an industry leader in chipsets and IP for 5G, which may help maintain a healthy stream of high-margin royalty over time.

Lastly, Qualcomm's focus on energy-efficient AI chips for data centers and automotive applications is set to expand its technology reach, and perhaps enhance its competitive advantage in the near future.