SOTS 2nd Hour: Qualcomm News, A Possible Trade Truce, & The Case Against AI Super-Intelligence 10/27/25

SPEAKER_1 (0:02)

At Janice Henderson Investors, we believe working together is the way to work better. Like combining your portfolio plans and our in-depth strategy, your valued assets and our valuable insights, your vision and our mission. Working in harmony to seek the right investment opportunities. Janice Henderson Investors, investing in a brighter future together.

SPEAKER_2 (0:32)

Huge savings on Dell AI PCs with Intel Core Ultra processors are here. And they are newly designed to help you do more faster. They can generate code, edit images, multitask without lag, draft emails, summarize documents, create live translations, and even extend your battery life. That's the power of Dell AI with Intel Inside. Upgrade today by visiting dell.com/ai-pc.

Carl Quintanilla (1:01)

Good Monday morning. Welcome to Squawk on the Street. I'm Carl Quintanilla with Leslie Picker and David Faber live at Post 9 of the New York Stock Exchange. Sara Eisen is on assignment traveling to Saudi Arabia for the country's annual investment conference. Meantime, record highs across the board, building on Friday's records as well. Most sectors are green. Some defensives are on the back foot. Materials, health care, and staples are red. And yields are pretty steady. Today, the case against super intelligence.

We'll talk to a professor who signed a petition to ban advanced AI, says unchecked advancements are like plain Russian roulette with every person on earth. Chip stocks jumping this morning as optimism builds around a China trade deal. Analyst Stacey Raskin will join us to break down the companies that might benefit the most from a better relationship between Washington and Beijing. Speaking of the chips, we got some breaking news this morning out of Qualcomm. And for that, we'll turn to Christina Parts-Nevalos. Good morning, Christina.

Christina Parthenopaeus (1:53)

Good morning, Carl. Well, Qualcomm just announced it's taking on NVIDIA in AI chips, a massive pivot for a company that built its empire on smartphones. So they're launching new data center AI chips starting in 2026 And they also just announced their first major customer, that would be Saudi backed AI startup Humane, targeting roughly 200 megawatts of capacity starting in 2026 So they're not essentially going after AI training. That's the market that made NVIDIA worth over $4 trillion. Qualcomm is targeting inference. That's running AI models, not necessarily building them. So every time you use ChatGPT or generate an AI image, that's inference. So training builds the models, inference uses them billions of times a day. Qualcomm thinks the market is going to be enormous. Nearly 6.7 trillion dollars in data center spending is expected through 2030 according to McKinsey. And so just capturing about 5 to 10%, maybe just even 5% of that market would transform Qualcomm's business. Their products are complete server systems based on their hexagon NPUs, so that would be neural processing unit chips, already used in billions of smartphones around the world. One key advantage, and this is where it gets a little technical, it's 768 gigabytes of memory per card. They're saying that's more than what NVIDIA and AMD offer in this particular rack type setting. That matters though for running larger AI models. And so this is where Qualcomm hopes to compete. They'll sell these complete systems that you're seeing on your screen right now for or also individual components. Qualcomm says even NVIDIA or AMD could become customers for some of their parts. But the timing really matters because, for example, OpenAI recently announced it's buying chips from AMD, showing big AI companies want alternatives to NVIDIA, especially within the inference market. But Qualcomm only named one customer so far. There's no pricing details. They're launching in 2026, which is years behind NVIDIA's dominance. But there's no taking away from the fact that this is a big new market for Qualcomm, guys.

SPEAKER_7 (3:48)

Yeah.

David Faber (3:48)

And I mean, as you point out, Christina, no shortage of potential competitors, obviously AMD and NVIDIA, but others as well. And we sometimes I only we talk enough about the fact that Alphabet, I mean, their TPU is for inference as well, right?

SPEAKER_7 (4:02)

Right.

David Faber (4:02)

And gives them potentially a leg up to a certain extent. This is kind of a different subject, but on a cost per token basis in terms of what it costs them. And a lot of this goes back to that, I guess, for the hyperscalers. What is your, you know, what's the cost per token going to be? I don't know where this thing is going to get priced and what it's going to be able to do, but that will become important.

Christina Parthenopaeus (4:20)

Yeah, normally in the chip world, they refer to it as total cost of ownership, TCO. And so Qualcomm didn't provide specific details. A lot of companies, how they benchmark it is different. They're saying that the total cost of ownership will be cheaper because of the power efficiency of said chips, because these aren't GPUs that are using a lot more power. The NPUs and more custom made chips tend to use a lot less. So that's going to be their selling point going forward. I'm sure in the coming months, they'll reveal some more data around that. But to your point, Google really stands out. Even Jensen Wong from NVIDIA has spoken many times about just how advanced Google has become in the TPU market and just they've been here for a long time. This isn't a company that just made a chip, you know, five years ago. So that's another company to look out for, no doubt.