Artificial intelligence is not just changing the game — it is the game. That was the message in the second day’s keynote at the Dell Technologies World conference in Las Vegas. It also featured a stark warning for partners and customers: The players who hesitate may not get a second chance to catch up.
“This is the most disruptive technology that I’ve seen in my career,” said Jeff Clarke, vice chairman and COO of Dell Technologies. “The threat is existential if you don’t respond.”
For MSPs integrating infrastructure for SMB clients, there’s a clear signal: Dell sees disaggregated infrastructure for AI as the next frontier. As a result, Dell is not just building tools for it; it is building the playbook.
From Token Factories to Real-world ROI
Dell Technologies World’s Day 2 keynote began with a barrage of AI momentum metrics, projecting $400 billion in AI capital expenditures this year alone. To put that in perspective, that level of investment outpaces the inflation-adjusted cost of the Apollo moon program.
A key driver of this explosion: The rise of “token factories.” According to Clarke, inferencing and reasoning models are 100 times more compute intensive than anticipated a year ago. “In 2024, 25 trillion tokens were generated. By 2028, that number will be 35,000 trillion tokens.” That’s three orders of magnitude increase over three years, which equates to an enormous demand for compute.
Jeff Clarke of Dell Technologies
For service providers, the real opportunity lies in delivering meaningful ROI to clients. Dell walked through its own AI journey as a proof point. Internally, the company deployed GenAI-powered service assistants across its global support operation. Built using Dell infrastructure and ran entirely on-prem, it delivered impressive outcomes: fewer dispatches, faster case resolution, and most importantly, happier customers.
“First, our agents close more cases and they do it faster,” said Clarke. “Our dispatch rates are down, our repeat dispatch rates are down even more … and customer satisfaction is up.”
To drive home the value-prop, Dell’s deployment used a single rack, required no additional power or cooling, and achieved ROI in under three months.
AI at the Edge — and in the Closet
One of the more important shifts is Dell’s firm commitment to meeting AI where the data lives. That means it’s not just in the cloud, but at the edge, in the rack, and on the client itself. Chairman and CEO Michael Dell laid out how Dell is equipping businesses of all sizes to “bring AI to the data” during the event’s opening keynote.
Arthur Lewis, president of Dell Technologies’ Infrastructure Solutions Group, unveiled new disaggregated infrastructure innovations under the Dell AI Factory banner, including the Dell AI Data Platform, Project Lightning, and PowerProtect Data Domain All-Flash.
“We’re redefining the future architecture of the intelligent enterprise,” Lewis emphasized. “To take advantage of the latest innovations in AI requires a unified data approach across a diverse set of sources, tools, and formats.”
Enterprise-grade AI capabilities are no longer exclusive to hyperscale deployments, Lewis explained. With optimized inferencing platforms, next-gen object storage, and modular, software-defined solutions like Dell Private Cloud, its partners and customers can now deliver AI infrastructure in environments as compact as a single server closet or as expansive as a multisite edge deployment.
The announcement of Dell Private Cloud and the underlying Dell Automation Platform is a big deal for channel partners. Built to simplify multi-hypervisor, disaggregated infrastructure, Dell Private Cloud allows customers to deploy validated stacks from vendors like Broadcom, Nutanix, and Red Hat in 90% fewer steps than manual processes. This delivers a private cloud cluster in just two and a half hours.
Dell is increasingly acknowledging that AI doesn’t just belong in the cloud. Dell built this private cloud strategy around flexibility, lifecycle efficiency, and empowering partners to offer a compelling on-prem alternative that meets modern AI demands.
“Managing your data effectively is the backbone of successful artificial intelligence,” said Lewis. “And now is the moment — the moment to lead, the moment to build, and the moment to shape what’s next.”
The Edge Isn’t Optional Anymore, and Will Include the Client
The proliferation of data at the edge (POS systems, IoT sensors, local apps, etc.) creates an opportunity to deliver managed AI services on-site, without the latency or privacy trade-offs of cloud-based inference. Devices and systems at the edge will generate about 75% of data, according to Sam Burd, president of Dell’s Client Solutions Group. Burd noted that Dell AI Factory has been designed for AI to run where it makes sense.
Sam Burd of Dell’s Client Solutions Group
That edge will increasingly include the client. “That data lives on the PC where the content is personal and meaningful, where insights need to be tailored and outcomes must be immediate,” he said. “That makes the PC the ultimate edge device. It’s your data, your edge, your way. Now fast forward to 2030 or a few years beyond that, and there will be 2 billion AI-PCs each capable of hundreds or thousands of TOPS.”
The fast-approaching end of life for Windows 10 was mentioned several times, with Dell urging the importance of refreshing PCs with AI capabilities as a means to future-proof customers for a quickly evolving AI future.
Agentic AI: Your New Team of Digital Coworkers
Clarke further leaned into the promise of Agentic AI. Dell defines it as autonomous software systems capable of reasoning, learning, and carrying out goals with minimal input. “These agents will quickly become autonomous,” he explained. “And autonomous agents working together as a team is what we call Agentic AI.”
That concept should excite MSPs. AI coordinates automated diagnostics, proactive remediation, and even business process automation. This is all aligned with modern MSP service delivery and designed to act as an extension of human teams. Dell’s demo of this approach in action — using AI agents to detect, escalate, and resolve infrastructure issues across a hybrid PC-data center environment — showcased how transformative agentic AI will be.
In fact, Dell projects that autonomous agents will complete one-third of all generative AI interactions by 2028, according to Clarke.
Partner Opportunities Are Front and Center
Throughout the keynote, Dell emphasized its ecosystem of partners. It name-dropped key integrations with NVIDIA, AMD, Intel, Google, Cohere, Hugging Face, Red Hat, and others. This made it clear that solution providers are central to bringing these technologies into the mainstream.
Arthur Lewis of Dell Technologies’ Infrastructure Solutions Group
“As you can see, we’ve been incredibly busy engineering the future,” said Lewis. “I can tell you that AI revolution is not a spectator sport … Let’s build what’s next together.”
You’re Probably Behind — and That’s Not Good
Dell challenged partners during its Day 2 keynote: Bring AI to your customers, or risk falling behind.
Don’t wait any longer for the right time to pitch AI. Whether as managed services, storage modernization, or productivity enhancement, now is the time. The technology is ready, and so are your customers, Clark said. “If you haven’t started, you’re behind. And if you’re not moving fast enough, you’re behind.”
In other words, get busy — or get beat.
➡️ Highlights from Day 1 of Dell Technologies World 2025.
