By Tae Kim

This article is from the free weekly Barron's Tech email newsletter. Sign up here to get it delivered directly to your inbox.

AI Enabler. Hi everyone. When Nvidia CEO Jensen Huang gave his two-hour keynote at last week's GTC AI event, one of the company's partners got a particularly prominent mention: Dell Technologies.

Huang emphasized his company's close relationship by presenting a Dell marketing slide during that said Dell has the "industry's broadest end-to-end Nvidia AI enterprise infrastructure" product line. Huang didn't do that with any other vendor.

Dell has become a leading provider of computers and servers for AI. In less than two years, the company built its AI server business to $10 billion in revenue. In its latest quarter, Dell projected 50% AI sales growth for 2025. Thanks to Dell's global distribution and services footprint, there may be no AI hardware maker better positioned to sell to enterprises and governments around the world.

In the AI server market, the company mainly competes with Super Micro Computer and Hewlett Packard Enterprise.

CoreWeave, a cloud-computing company that runs on graphics-processing units, and Elon Musk's xAI both are using Dell for their AI infrastructure buildouts. Dell recently won what's reported to be a large multibillion-dollar AI server order from xAI.

Barron's Tech recently spoke with Dell founder and CEO Michael Dell to discuss the future of AI, the company's relationship with Nvidia, and the state of AI demand from governments around the world.

Here are edited highlights from our conversation with Dell:

Barron's: Compared with the PC revolution and the transition to mobile, what's your view of how big this AI computing shift will be?

Dell: It's every bit as big as the ones you mentioned. The historical context is that for the past 50 years, the technology and computing world was all about calculating and computing. Now we've shifted into machines that help us think and are thinking for us. What are these models creating? They're creating intelligence.

It is very different from creating calculations. How big is the market for intelligence? It's very big. Every company that we work with wants to figure out how to utilize these tools to make their business more efficient, to make it more productive, to grow, to expand, to develop products faster, to unleash the power of their data, and to be able to innovate more effectively. We're still at the early stages of this.

If AI were a baseball game, what inning are we in?

I always think of what we do as a race with no finish line. It's not like there's ever an end to the game, right? The models are amazing. The computing power is amazing, but the models are the worst they'll ever be. A few years from now, they're going to be way more incredible.

Last year, we've helped about 2,200 of our largest customers create these Dell AI factories. I think if we meet in a couple of years, we will have helped a million customers. It won't just be the large enterprises, it'll be the medium-sized companies, the small businesses. If you think about it as S-curve adoption, we're still in the first single-digit percentage of the S-curve.

Can you specifically talk about how you won this latest deal with xAI?

I won't speak for the customer. But I think they looked at our ability to deliver, our engineering, our supply chain and how we performed on the floor. We'd already delivered a lot of systems to them. They worked really well.

We're a reliable, predictable company. If we say we're going to deliver 100,000 of these things on June 1st, we deliver 100,000 on June 1st or before. We're not going to tell them we can do it and then not do it. That's just our culture.

What kind of AI work are your customers doing?

Customers are rapidly moving from tests and proof of concepts into deployment. We've seen roughly two-thirds of companies have successfully made that transition into deployment. It's everything from enhancing the services that organizations deliver to their customers to coding assistants, along with content development and content creation.

There are tons of uses of this computing power that are outside of large language models. We've got quantitative traders and all sorts of people doing clever things with machine learning and deep learning and tapping into this power in new and interesting ways. We created this past year over 100 solutions. I think of these as reference platforms and designs that help catalyze the activity inside every vertical sector.

How close is your relationship with Nvidia?

I've known Jensen for 31 years. We were reminiscing about the [Nvidia] Riva 128 which came out in 1997. That was kind of a pivotal moment.

When writing my book, an Nvidia executive told me he had to call you to let you know Dell's allocation would be 50% lower due to Riva 128 manufacturing problems, which was a difficult moment because this was Nvidia's first successful product.

I remember that. Those challenges create incredible partnerships and friendships, usually. [laughter]. Pressure makes diamonds. We grew up together building PCs and workstations in the early days.

We provide computing infrastructure to more companies and governments than any other company. Nvidia appreciates that. They want to get their technology in the hands of all the customers that work with Dell.

Nvidia recently showed off their GPU road map, showing one AI server will have 576 GPUs in two years -- four times more than today's model. What will customers do with all this power years from now?

Good question. I don't think anyone really knows. There have been these cascading advances that keep coming. Now, we have test time compute, incredible deep reasoning. You play around with these models and it's like, my God, I got a PhD thesis on this complicated question I asked. It's incredible.

We're still at the beginning in servicing and enabling human potential in an extraordinary way. If every person on the planet practically has relatively easy access to these tools, it's a bright future for humanity. I think we'll see tremendous acceleration of scientific advancement in drug discovery, in health, and energy and just about every domain.

What are the biggest, most unsolved problems, scientific questions out there in the world? Those will get knocked down at a faster and faster rate.

Sovereign AI has become an increasingly big topic. What are you seeing right now in terms of AI demand from governments around the world? Are you optimistic about this year?

Absolutely. I travel around the world physically and virtually. We have teams in 180 countries, and we see just about every country that has some level of resources, they want to have their own AI models that represent their language, their culture, their beliefs, and their data sovereignty. They prefer to have their own. There's definitely a big trend there.

Thanks for your time, Michael.

This Week in Barron's Tech

   -- CoreWeave's IPO Looks Pricey. History Offers a Different View. 
 
   -- Apple AI Upgrade Won't Come This Year, UBS Says 
 
   -- Alibaba Stock Drops. Its Chairman Sees an AI Bubble 
 
   -- Mobileye Stock Gains. Why a New Deal Has Shares Soaring. 
 
   -- Cloudflare Stock Gets a Double Upgrade 

Write to Tae Kim at tae.kim@barrons.com or follow him on X at @firstadopter.

This content was created by Barron's, which is operated by Dow Jones & Co. Barron's is published independently from Dow Jones Newswires and The Wall Street Journal.

(END) Dow Jones Newswires

March 26, 2025 16:04 ET (20:04 GMT)

Copyright (c) 2025 Dow Jones & Company, Inc.