Breaking: AI Demand Explodes, Dell and NVIDIA Reveal New Infrastructure
NVIDIA CEO Jensen Huang declared Monday that demand for AI is entering a 'parabolic' phase, as Dell Technologies World in Las Vegas showcased a new generation of enterprise AI infrastructure. The announcement, delivered alongside Dell CEO Michael Dell, signals that the era of experimental AI pilots is over, replaced by large-scale agentic AI deployments.
"We’ve now arrived at the era of useful AI, which is the reason why demand is going parabolic, utterly parabolic," Huang told the audience. He emphasized productivity gains that are compressing timelines: "What took months now takes weeks. What took weeks now takes days. And what takes days now takes hours."
Dell set the context, projecting that worldwide AI infrastructure spending could reach $3 to $4 trillion by 2030, with token consumption forecast to surge 3,400% over the same period. "There is a massive AI investment boom that's already underway, and a productivity boom is beginning," Dell said. "The rate of change has gone parabolic, and it's not slowing down."
Key Products: Vera Rubin NVL72 and the New Dell AI Factory
At the center of the news is the Dell PowerEdge XE9812, built on NVIDIA's Vera Rubin NVL72 platform. Dell claims this system delivers up to 10x lower cost-per-token for large-scale agentic AI inferencing compared to the previous NVIDIA Blackwell architecture. The XE9812 is joined by three new servers—PowerEdge XE9880L, XE9885L, and XE9882L—the first Dell systems using NVIDIA HGX Rubin NVL8. These racks can support up to 144 GPUs per rack with 100% direct liquid cooling and up to 5.5x the performance of HGX B200.
Networking gets a major upgrade with the new Dell PowerSwitch portfolio, featuring NVIDIA Quantum-X800 InfiniBand with liquid-cooled, co-packaged optics and Spectrum-6 Ethernet. Dell also introduced Dell PowerRack, a fully integrated compute-networking-storage system engineered as a single unit, reducing integration overhead for enterprise-scale AI workloads. On the CPU side, the Dell PowerEdge M9822 and R9822 servers bring NVIDIA Vera CPUs, purpose-built for agentic AI data pipelines and sandboxed tools.
Dell cited early adopters including Lilly, Samsung, and Honeywell—5,000 enterprises now running AI workloads on Dell AI Factories with NVIDIA. These factories are designed to run frontier models and autonomous agents securely behind the enterprise perimeter.
Background: From Pilots to Production
The announcements mark a tipping point in enterprise AI adoption. Over the past two years, most companies limited AI to experimental pilots, but Dell and NVIDIA now see a shift toward production-scale inference and agentic AI. The cost-per-token reduction—10x with Vera Rubin—directly addresses the biggest barrier to deployment: operational expense. Jensen Huang characterized the moment as the start of a 'productivity boom,' driven by AI that is actually useful for business processes.
Michael Dell framed the investment trajectory: AI infrastructure spending hitting trillions by 2030, with token consumption growth outpacing any previous technology cycle. The Dells' combined hardware and software stack aims to provide enterprises a turnkey path from pilot to production without the complexity of assembling separate components.
What This Means: A New Era of Agentic AI at Scale
For enterprises, the immediate implication is that large-scale AI deployment is now economically feasible. The 10x cost-per-token improvement from Vera Rubin NVL72 means agentic AI—systems that can plan, execute multi-step tasks, and use external tools—can run at a fraction of previous costs. Dell's integrated PowerRack approach further lowers the barrier: IT teams no longer need to manually configure compute, storage, and networking for AI workloads.
The focus on agentic AI also signals that the industry is moving beyond chatbots and simple inference. Agents that can autonomously query enterprise data, run sandboxed code, and orchestrate workflows will become mainstream. Huang's comment about productivity leaps—from months to weeks to days—highlights the compounding effect of AI that acts, not just responds.
With 5,000 enterprises already using Dell AI Factories, the technology is moving from early adopters to early majority. The combination of dramatically lower cost-per-token, liquid-cooled high-density racks, and purpose-built CPUs positions Dell and NVIDIA to capture a significant share of the projected $3-4 trillion AI infrastructure market by 2030.