Dell Technologies World: A Bright and Beautiful Road Ahead

Dell Technologies World 2026 Day One keynote: Michael Dell and NVIDIA CEO Jensen Huang announced the Dell AI Factory with agentic AI capabilities, new liquid-cooled PowerEdge XE servers, Dell Exascale Storage, Dell Deskside Agentic AI for edge inference, expanded model availability including Google Distributed Cloud with Gemini 3.0, OpenAI Codex, SpaceXAI Grok and partnerships with Palantir, ServiceNow and Hugging Face.

Dell Technologies’ 2026 annual conference opened with Michael Dell declaring AI infrastructure — not cloud, not software — as the core enterprise differentiator, unveiling the Dell AI Data Platform with NVIDIA, new direct liquid-cooled compute, Dell PowerRack, edge AI workstations capable of running trillion-parameter models and a broad partner ecosystem spanning Eli Lilly and Company, Honeywell, Ascension, Samsung and NVIDIA.

Michael Dell always brings big ideas to the Dell Technologies World main stage. Monday, the ideas felt bigger, more immediate and more real.

In an opening keynote that took a packed house at the Venetian Las Vegas on a journey through the modern enterprise, the Dell Technologies chairman and CEO shared his conviction that the AI era had arrived in earnest.

“Abundant intelligence is here,” he said. “It’s not coming; it’s here right now.”

Intelligence is becoming infrastructure. Just as electricity transformed the world when it left the power plant, AI is transforming the world as it leaves the screen and enters the physical world. AI is in the ambulance, on the oil rig, on the factory floor and in orbit. Dell’s job, Michael said, is to build the distributed infrastructure that makes that possible.

“We are opening the door to a new era of progress across every field of human endeavor,” he said.

Since ChatGPT launched in late 2022, the industry has moved from a chatbot that could write a decent email to self-improving AI agents that write code, run workflows and operate around the clock. Coding, Michael noted, is now AI-native. The barrier between imagination and execution is collapsing.

That means AI is no longer a feature. It is becoming the operating model of the modern enterprise and the companies that redesign their work around it will gain advantage faster than any generation in history.

“AI can become the most concentrating technology in history, or the most democratizing,” Michael said. “We are choosing democratizing. Powerful AI wherever it’s needed. Open, secure and yours.”

From the boardroom to the factory floor

To bring the journey to life, Michael highlighted the modern boardrooms, manufacturing floors, data centers and hospitals where AI is moving from pilot to production.

The first stop was Lilly, a 150-year-old company where AI-powered solutions are compressing drug development timelines from years to weeks. Diogo Rau, Lilly executive vice president and chief information and digital officer, joined Michael on stage to discuss how digital twins are driving efficiencies and accelerating the delivery of life-saving products.

“Medicine isn’t just about discovering something new,” Rau said. “It only works if you get it out at scale.” With AI, Lilly achieves scale and speed Rau never imagined. “We’re solving problems that we weren’t thinking about ten years ago,” he said. “In 150 years, we may be able to end disease as we know it.”

Michael reflected, “What you just saw is not a chatbot. It’s intelligence in the physical world. It’s a faster path from science to medicine.”

The journey continued with a video look at Samsung Electronics, where AI is accelerating partnerships from proof of concept to production in modern semiconductor manufacturing.

Honeywell CTO Suresh Venkatarayalu then joined Michael to explore what AI looks like when it works with machines, sensors, supply chains, safety systems and real-time decision-making. Venkatarayalu said the 140-year-old company’s customers’ systems are fragmented. Honeywell optimizes their operations and with AI has the ability to “contextualize assets and processes much better in order to optimize operations.”

The agentic shift

If the first wave of AI improved workflows and delivered meaningful productivity gains, Michael made it clear that a more consequential wave is now underway. The industry is moving toward autonomous agents that plan, reason, execute, adapt and close the loop.

“Agents are not something you just simply bolt onto legacy systems,” he said. “They are digital workers. They have memory and credentials and access and the ability to take action.”

The implications are significant. The companies that get there first will rapidly distance themselves from the rest. Those that don’t risk being left behind.

That shift also reshapes the security landscape. Agents have credentials, access and autonomy, which means security can no longer be designed only around human users. It must govern non-human workers operating at machine speed. Dell’s answer is visibility and control at every layer — zero trust principles built into AI infrastructure from endpoint to core, with Dell Trusted Device providing visibility into the PC and ControlVault securing credentials so that only authorized identities, human or machine, can access customer systems. “You can’t protect what you can’t see and you can’t manage,” Michael said.

Your data is your differentiator

Everyone, Michael pointed out, has access to the same foundation models. What separates companies is their data — their unique, proprietary, hard-won business knowledge. Agentic AI is only as good as the data it can trust, access and act on. If that data is siloed, agents are blind.

Dell announced the Dell AI Data Platform with NVIDIA to bridge the AI execution gap, not just the ambition gap. Its Data Orchestration Engine unifies structured and unstructured data across environments, delivering 12x faster vector indexing, 6x faster data querying and 19x faster time to first token. An enhanced Analytics Engine, built with Starburst, adds 3x faster SQL on NVIDIA Vera CPUs and 6x faster SQL on Blackwell GPUs. Bank of America is among the early customers using the Starburst, NVIDIA and Dell combination for AI-driven analytics under strict governance and regulatory requirements.

In storage, Dell introduced Dell Exascale Storage, a unified rack architecture supporting PowerScale, ObjectScale, PowerFlex and the Lightning file system, capable of up to six terabytes of throughput per second per rack. The Lightning File System, PowerScale and ObjectScale are in production with customers including Coreweave, Nscale, IREN, Fluidstack, Boostrun, McLaren, Hudson River Trading, Quadrature, 2Sigma and Optiver.

Networking additions include the PowerSwitch Spectrum-6 and the NVIDIA Quantum-X800 with liquid-cooled, co-packaged optics. At the compute layer, the PowerEdge XE9812, built on NVIDIA Vera Rubin NVL72, delivers 10x lower cost-per-token than Blackwell. A new lineup of PowerEdge XE servers on NVIDIA HGX Rubin NVL8 supports up to 144 GPUs per rack with fully direct liquid-cooled compute nodes, delivering 5.5x more power than B200 GPUs. Dell PowerRack packages compute, networking and storage as a validated, turnkey rack-scale unit.

AI at the edge, and a conversation with Jensen

Agentic AI is here and it runs on-prem. Dell announced Dell Deskside Agentic AI, Dell Pro Precision workstations with NVIDIA NemoClaw and Dell Services, for building, testing and fine-tuning agents locally on models ranging from 70B to one trillion parameters, without unpredictable cloud costs or IP exposure. NVIDIA OpenShell is now supported across the Dell AI Factory for agent development from workstation to server, and Dell support for NVIDIA AI-Q 2.0 blueprints provides a foundation for multi-agent workflows.

Dell’s ecosystem is second-to-none. The company announced an expanded Dell Enterprise Hub on Hugging Face with MiniMax-M2.7, DeepSeek Pro, DeepSeek-V4, GLM 5.1 and Kimi K2.6, joining existing models including Gemma 4, NVIDIA Nemotron Super 3, Mistral Small 4 and Arcee Trinity-Large-Thinking and Mistral AI. Google Distributed Cloud with Gemini 3.0 is now available on PowerEdge XE9780 servers in a confidential computing environment. Dell and OpenAI are building an on-premises solution based on OpenAI Codex, connecting GPT and GPT-Codex models via the Dell AI Data Platform. SpaceXAI Grok is available as an on-premises or hybrid, secure enterprise-grade AI assistant. Reflection AI’s open-source models are coming on-premises for regulated industries and sovereign deployments. Palantir Foundry and AIP are coming on-premises via ObjectScale and PowerFlex. ServiceNow customers will be able to run Otto on the Dell AI Factory with NVIDIA.

NVIDIA Founder, CEO and President Jensen Huang joined Michael on stage for a conversation covering the agentic shift, the value of enterprise data and the economics of on-premises AI infrastructure.

“We now have, for the very first time, useful AI,” Huang said. “That’s why demand is going parabolic. Everybody is using agents all over the place. We’ve now arrived at the era of useful AI. Our company has always gone fast. It’s really going fast now. Our ambition has changed.”

Huang’s “enthusiasm is infectious,” Michael said afterward, as is his “passion for innovation and his belief that accelerated computing can bring the power of AI to the world.”

Doing it responsibly, at scale

Compute costs per token are dropping 80 percent annually while token consumption is projected to grow 3,400% by 2030. By then, worldwide AI infrastructure spending could reach $3 to $4 trillion per year. More than 5,000 enterprise customers are running AI workloads on Dell AI Factories and Dell is able to deliver hundreds of racks per week, operational within hours.

On power and cooling, Dell’s PowerCool Enclosed Rear Door Heat Exchanger reduces cooling energy consumption by up to 60% and cuts annual energy costs by up to one-third. The PowerCool CDU delivers more than 220 kilowatts of cooling capacity in the smallest form factor capable of supporting the Vera Rubin NVL72. Michael noted that even as total energy demand grows with scale, the energy required per token continues to fall.

The real story begins now

The keynote closed with a video look at Ascension, one of the U.S.’s leading nonprofit health systems, where Michael said, intelligence is measured not in tokens but in time, accuracy, compassion and lives saved.

AI is helping Ascension with basic things like chest X-rays and reading charts and test results. But it’s also helping surgeons pioneer cutting-edge techniques. The organization’s vision is that “technology meets the needs of mankind every single minute of every single day,” said cardiac surgeon Dr. Chuck Frazier.

It was a fitting end to a morning that kept returning to the same essential point: technology matters most when it starts with the outcome: A better product. A safer process. A breakthrough. A cure. A healthy child.

“For too long, the AI conversation has been trapped inside the screen,” Michael said. “The real story begins now. AI is moving into hospitals, factories, schools, energy grids, laboratories, cities, homes and even into orbit, solving problems at the scale of humanity. The road ahead is bright and it is beautiful,” he said, “and we will be with you on it every step of the way.”