bytevyte
bytevyte
Language
EN DE ES FR PT IT
ai-beats —

Dell and NVIDIA Launch AI Factory Optimized for Agentic AI Inference

agentic AI inference

Dell Technologies and NVIDIA have expanded their collaborative AI infrastructure with the launch of the Dell AI Factory, a platform specifically engineered to support the deployment of autonomous agents at scale. The centerpiece of this announcement is the Dell PowerEdge XE9812 server, which integrates the new NVIDIA Vera Rubin NVL72 platform. This hardware combination is designed to handle the complex orchestration and long-context management required for agentic AI inference, a growing segment of the enterprise market.

The Dell AI Factory aims to reduce the financial barriers to deploying sophisticated AI agents. According to data released by NVIDIA this week, the new systems provide up to a 10x reduction in cost per token for agentic AI inference when compared to previous Blackwell-based architectures. This efficiency gain is paired with significant performance improvements; agent sandboxes operate 50% faster on the Vera platform than on traditional x86 CPUs, while enterprise data queries see a 3x speed increase. These metrics suggest a shift toward hardware optimized for the specific logic and tool-calling patterns of autonomous systems.

Strategic Infrastructure for Agentic AI Inference

At the heart of the new server is the NVIDIA Vera CPU, which features 88 custom-designed Olympus cores. This processor is NVIDIA's first CPU built specifically for the orchestration tasks inherent in agentic workflows. With a memory bandwidth of 1.2 TB/s per socket, the Vera CPU addresses the data throughput bottlenecks that often hinder autonomous agents. Initial units of the Vera CPU have already been delivered to major AI laboratories, including OpenAI, Anthropic, and SpaceXAI, as well as Oracle Cloud Infrastructure.

The partnership also introduces liquid-cooled rack-scale systems capable of supporting up to 144 GPUs per rack. This high-density configuration utilizes NVIDIA Quantum-X800 InfiniBand and Spectrum-6 Ethernet networking to maintain the low latency required for real-time agent interactions. Beyond hardware, the collaboration extends to software and cloud integration. Dell and NVIDIA are working with Google to bring Google Distributed Cloud (GDC) with Gemini 3.0 to Dell hardware, allowing enterprises to deploy frontier models within their own secure on-premises environments.

Enterprise adoption of these integrated solutions is already substantial, with over 5,000 organizations currently utilizing Dell AI Factory infrastructure. Companies such as Eli Lilly and Samsung are among the early adopters leveraging these systems for large-scale AI workloads. By providing a pre-integrated stack that combines NVIDIA's specialized silicon with Dell's enterprise hardware, the two companies are positioning themselves as the primary providers for the next phase of corporate AI deployment, where the focus moves from simple chatbots to autonomous agents capable of executing complex business processes.

While we strive for accuracy, bytevyte can make mistakes. Users are advised to verify all information independently. We accept no liability for errors or omissions.

Sources

NVIDIA CEO Jensen Huang at Dell Technologies World: ‘Demand Is Going Parabolic, Utterly Parabolic’

Vera Arrives: NVIDIA’s First CPU Built for Agents Lands at Top AI Labs

AI-generated image.

✔Human Verified

Share