Artificial intelligence adoption has brought hardware back in style, with Dell Technologies Inc. having an astounding week.

The hardware firm saw an 88% jump in revenue and its stock closed up at 33% last week. These amazing numbers reflect how AI is pushing enterprises to seek more compute power — and how Dell is taking advantage of the opportunity.

“In the old [personal computer] revolution, the PC set the agenda, then the servers came on,” said John Furrier (pictured left), executive analyst for theCUBE Research. “Servers came after PCs. If you look at the AI business, it’s completely inverted. The servers are coming first and the PCs are coming second. Dell has engineered essentially a super server. They’re essentially shipping super servers that are all configured.”

On the latest episode of theCUBE Pod, Furrier and Dave Vellante (pictured), chief analyst for theCUBE, Research discussed Dell’s incredible week and what it means for the AI infrastructure boom. They also looked at Snowflake Inc. ahead of the upcoming Snowflake Summit and how Anthropic is faring against OpenAI.

Dell sees huge growth off of AI hardware business

Dell’s bet that AI adopters will gravitate toward a hybrid model has so far paid off. Dell’s revenues were almost 44 billion for the quarter, and that’s mostly due to its AI servers. Since Dell operates with such low operating margins, the company currently has very little competition in the PC space, according to Vellante and Furrier.

“Dell’s rack scale systems have such a high GPU utilization,” Furrier said. “It’s the preferred purchase for all these build-out infrastructure players like the neoclouds. Everyone else has cobbled together the networking, the storage, the interconnects … when Dell’s just shipping a prefabricated rack, six hours from shipping it to install[ation]. That’s essentially just turning on the switch.”

Companies specializing in hardware for AI saw big gains across the board. Even though data storage company NetApp didn’t show quite as impressive numbers as Dell, it still saw its stock rise more than 30% in recent weeks. Time will tell if the momentum around AI infrastructure will continue or if new obstacles will arise.

“The question is: Is it sustainable?” Vellante said. “I do think it’s a lumpy business. If the hyperscalers and the capex slow down a little bit and maybe the CoreWeaves where Dell’s doing a lot of business and the AI clouds aren’t spending as much, that will have an impact. Enterprises are air-cooled, and the question is: Will they eventually refactor their data centers to be water cooled?”

Furrier predicts that as the industry shifts toward distributed AI computing, water-cooling will only be necessary for the big factories. He sees the next big obstacle to AI adoption as GPU utilization, since the enterprise world is spending tons of money on capital expenditures without actually getting the most out of their AI infrastructure.

“If you look at most of the neoclouds, this is a huge problem that nobody’s talking about,” Furrier said. “A lot of these GPUs are underutilizing except for Nvidia. You had Arthur Patterson basically say to theCUBE, ‘99% utilization with the Dell rack scale servers.’ That’s an incredible number. Most of the self-built infrastructure has between 30 and 60% utilization of GPUs.”

Agentic AI moment recalls early PC adoption

Another winner this week was Anthropic, which raised $65 million in new funding — more than OpenAI. Snowflake also made a strong showing; its stock jumped 36% Thursday, in part because Snowflake committed to buy more compute from Amazon Web Services.

TheCUBE’s coverage of Snowflake Summit this week will give insight into where the data giant is headed in the AI era.

“Right now the simplicity of doing things like Cortex Code and Snowflake Intelligence — and it’s all governed and all inside that lovely blue blanket of Snowflake that’s governed, that’s shareable, that’s secure — is paying off for the company,” Vellante said. “Databricks is not a public company, but I would suspect you’re seeing a similar dynamic play out there. I think you’re going to see a lot of interesting innovations and advancements at Snowflake Summit this year.”

In a recent analysis, Vellante observed that the AI era echoes the PC era, with individuals employing AI models and agents in their work the same way that people used spreadsheets and word processors. As individuals and organizations adopt different agents, the result could be islands of intelligence.

“While the CEOs initiated a top-down mandate for AI, a lot of the productivity is happening at personal levels,” Vellante said. “People downloading NemoClaws and Airmees, doing their own thing and driving productivity and paying for it themselves. And it’s like the PC era … obviously, some differences, but that’s how it happened before you had connected servers and networks.”

Watch the full podcast below to find out why these industry pros were mentioned:

Chirantan “CJ” Desai, president and CEO at MongoDB
Brian J. Baumann, founder of NYSE Wired and director of capital markets, technology at NYSE
Vladimir Putin, president of Russia
Donald Trump Jr., American businessman
Michael Dell, chairman and CEO at Dell Technologies
Jensen Huang, president, co-founder and CEO at Nvidia
Jeff Clarke, chief operating officer and vice chairman of Dell Technologies
Tom Sweet, former CFO at Dell Technologies
David Floyer, analyst emeritus at theCUBE Research
Arthur Lewis, president for infrastructure solutions group at Dell Technologies
Chen Goldberg, senior vice president of engineering at CoreWeave
George Gilbert, principal analyst at theCUBE Research

Here’s the full episode of this week’s theCUBE Pod:

You can also listen to us on Apple Podcasts or on Spotify. And for those who prefer to watch, check out our YouTube playlist. Tune in now, and be part of the ongoing conversation.

Photo: SiliconANGLE