Takeaways From CES: Nvidia, AMD Updates That Could Move Our AI Names

We are roughly one day into CES 2026, and as expected, the wave of product introduction announcements and keynote catalysts from Nvidia (NVDA) and AMD (AMD) are serving to move AI and other tech stocks.

The event is a mixture of "here’s what’s coming later this year" and bigger picture “here’s the future” comments. Looking at Nvidia’s keynote address, its Rubin chipset platform that will ship in 2H 2026 is an example of the former, while Alpamayo, a family of open models built for autonomous driving, is more of the latter.

Our take on Nvidia’s Rubin is that it extends the firm's lead in the AI chip market, but we would not be surprised to see margin issues as that new chipset is launched and ramps, just like we saw with Blackwell last year.

In terms of Alpamayo, we’ll reiterate our recent comments on robotaxis and autonomous driving, adding that we’ll want to watch the regulatory environment as well as adoption by both auto OEMs and consumers. Following what is shaping up to be the collapse of the domestic EV market following tax subsidies expiring last year, consumer adoption of autonomous vehicles will be a key item to track, as will fleet adoption from the likes of Hertz (HTZ) and Uber (UBER) .

While we won’t ignore Nvidia’s comments on the autonomous market or the ones it made about robotics, let’s remember the company’s Automotive and Robotics segment accounts for just 1% of revenue. While that is poised to grow in the coming quarters, so too is the bread-and-butter Data Center business. That means at least in the next several quarters, Nvidia’s business and its shares will be driven by AI and data center developments. Fortunately, that outlook remains more than favorable.

Another future-facing comment came from AMD’s Lisa Su, who shared that the company sees AI going from 1 billion active users to 5 billion within five years. That raised our eyebrows and most likely a few others. In the past, we’ve shared that we don’t get overly hung up on the precise numbers when figures like these are shared, but focus more on the vector and the velocity. In this case, those are “higher” and “fast.” That meshes with our thinking on AI adoption and usage across the enterprise, with consumers and other entities like federal, state and local governments.

As smart as Su is, we’re not going to simply take her word for that expected ramp in AI active users. We’ll continue to follow the data. But if Su’s math is only 80% correct, the level of users combined with the growing number of use cases for AI, well, that’s the digital infrastructure capacity issue we’ve been talking about with you when it comes to several Portfolio holdings.

AMD, Intel, AI PCs and Qualcomm

One of the pitfalls that we’ll look to avoid is reading each announcement in a vacuum rather than contemplating what is being announced across the competitive landscape. For example, Su unveiled the company's upcoming AI data center platform, but she also introduced a new line of AI processors as the company thinks AI-powered personal computers “are the way of the future.”

Su wasn’t alone in targeting the AI PC market — Intel (INTC) also announced its Panther Lake AI chip, which targets the AI PC market and the one for handheld gaming devices. Now, it’s true that Intel has been an also-ran of late, especially in the data center market, but that also means we could see Intel lean on price to win business against AMD and Qualcomm (QCOM) . And yes, Qualcomm also announced its new AI PC offering, Snapdragon X2 Plus, that will start shipping in the coming months.

Here’s the thing: the AI PC market is a key part of Qualcomm’s revenue diversification strategy. Greater competition in that market will be something to watch as it relates to the speed of that diversification, but also the margins tied to it. We’ll add that to our growing list of concerns as it relates to QCOM shares, even as they trade higher on Tuesday.

On the positive for Qualcomm, at CES it announced its move into the robotics market with chips for industrial robotics as well as “advanced full-sized humanoids." Alongside that announcement, Qualcomm shares that it is working with the likes of Advantech, APLUX, AutoCore, Booster, Figure, Kuka Robotics, Robotec.ai and VinMotion to bring deployment-ready robotics at scale. Qualcomm also announced an expanded relationship with Google (GOOGL) for the automotive market as well as a new chipset offering for the IoT market.

The sobering remark we’ll make is that the products using the chips from these companies will hit shelves in the coming months, and from there, it’s a question of consumer demand. The point we’re getting at is, if memory supply and subsequent costs turn out to be a greater headwind to PC, smartphone and other connected devices shipments, design wins aren’t going to count for much if folks aren’t buying the products.

With that in mind, our next known catalyst to watch will be quarterly results next week from Taiwan Semiconductor (TSM) and the forward-facing comments it makes for its various end markets. Foxconn already shared that it sees seasonal weakness unfolding outside of its AI and data center server business.

More Pro Portfolio

At the time of publication, TheStreet Pro Portfolio was long NVDA, QCOM and GOOGL.