The Nvidia DSX Pod is a preconfigured AI Factory Building block, cutting time to token production to weeks from months. NVIDIA
Nvidia announced a long list of new technology and ecosystem initiatives today at the annual GTC-Berlin event focussed on enabling Europe to emerge from their slumber and begin innovating and investing in AI. Reducing the effort and expertise needed to install and run an AI factory could help sovereign AI Factories as well as Enterprises jump start the European AI industry at a time of increasing skepticism about the reliability and sustainability of USA-based suppliers. (Note that Nvidia is a client of my firm, Cambrian-AI Research.)
Nvidia CEO Jensen Huang certainly believes in the region’s potential. “Every company in Europe will become an AI company,” Huang said at Nvidia’s GTC conference in Berlin. “No region has dedicated itself to making mechatronics, electronics and mechanical systems at the level of precision and craft as the European region.”
Nvidia is putting its money where its mouth is, committing nearly $10 billion to Europe’s technology ecosystem. It has already established an early lead in supplying the processors and software needed for robotic ‘brains’, giving it a commercial interest in Europe’s industry succeeding.
The DSX Pod AI Factory
One of the issues slowing down AI expansion and growth (if one can say that) is that it typically takes six months to a year to provision the AI technologies, once the data center is built with power and cooling. Environmentally conscious Europeans will be please to note that the DSX Pod enables European companies to reduce water consumption for cooling by up to 100% and generate the same AI token output with up to 30% less power. And the DSX Pod factory can go from the truck to token product in as little as two weeks.
The DSX Pod AI Factory is roughly the size of a standard shipping container and weighs about 28,000 kilograms (nearly 62,000 pounds), Nvidia said. It’s designed to be shipped on a truck on normal roads and all the cabling and liquid cooling for the Pod are assembled and tested at the factory, the company said.
The DSX Pod is a reference design that will ship from Nvidia’s manufacturing partners. Nvidia said it was working with manufacturers including Foxconn, Gigabyte, Lenovo, HPE, Quanta and Pegatron. Nvidia expects deliveries to start in early 2027.
As AI factories scale, power becomes a constraint which makes it critical for every watt to produce more intelligence. Nscale is using the NVIDIA DSX Platform to coordinate power allocation with workload requirements. In tests using NVIDIA GB300 NVL72 systems, Nvidia partner Nscale demonstrated that NVIDIA DSX MaxLPS enabled 37% more GPUs and nearly 50% higher aggregate inference throughput within the same provisioned power budget.
Nvidia has already engaged with major delivery and implementation partners. This doesn’t look like marketing hype; major partners are already lined up to engage,
Schneider Electric has developed power and cooling reference designs for NVIDIA Vera Rubin infrastructure and is working with AVEVA on digital twins spanning infrastructure design, construction and operations.
Procore is helping manage AI factory builds with its construction management platform, integrated with DSX Sim.
Siemens is using DSX Sim to simulate power, cooling and acoustics, while Dassault Systèmes and Vertiv are developing digital twin models of Vertiv’s OneCore infrastructure, enabling simulation and optimization through model-based systems engineering.
ABB is developing SimReady assets with NVIDIA DSX Sim, enabling engineers to evaluate source-to-rack architectures before construction while advancing 800 VDC development through its newly announced Infinitus Direct Current portfolio.
As AI factories scale, power becomes a constraint which makes it critical for every watt to produce more intelligence. Nscale is using the NVIDIA DSX Platform to coordinate power allocation with workload requirements. In tests using NVIDIA GB300 NVL72 systems, Nscale demonstrated that NVIDIA DSX MaxLPS enabled 37% more GPUs and nearly 50% higher aggregate inference throughput within the same provisioned power budget.
Ensuring The Security of Agentic AI
Agentic AI has run into a wall of worry of late around the security risks that arise when AI is acting, not just searching. To ameliorate these real concerns, Nvidia announced a new security platform called the OpenShell Reference Architecture for Safe AI Agents. The platform runs on leading enterprise operating systems and container orchestration layers, including Linux and Kubernetes infrastructure powered by Canonical, SUSE and Red Hat, as well as Microsoft Windows using Microsoft Execution Containers (MXC). It can also run on cloud environments from all three major cloud services: Amazon Web Services, Google Cloud and Microsoft Azure.
Built on the foundation of Nvidia OpenShell secure runtime software, the architecture is designed to give organizations full-stack control to deploy AI agents with clear boundaries around what agents can and cannot access and do. Every agent has permissions, and run initially in a sandbox to allow users to validate the appropriate level of security.
NVIDIA-OpenShell-Reference-Architecture NVIDIA
Based on the extensive partner engagements, I believe that Nvidia has now cracked the Agentic AI security threats that recently received significant bad press after OpenAI agents hacked into Hugging Face’s infrastructure. Anthropic, AWS, Cisco, CrowdStrike, Dassault Systèmes, Dell Technologies, Google Cloud, HPE, IBM, Meta, Microsoft, OpenAI, Palo Alto Networks, Perplexity, SAP, ServiceNow, Siemens, SpaceX Super Intelligence are among those working with the OpenShell Reference Architecture to enhance the security of agent deployments.
Implementation partners include Aleria, Baseten, Cassava, CoreWeave, Crusoe, DigitalOcean, Firebird, Fireworks, Firmus, G42, GMI Cloud, IREN, Lambda, Lightning AI, Nebius, Nscale, Radiant Verda and Vultr are supporting the OpenShell Reference Architecture to enable secure agentic deployments for their customers.
GTC Is Not A Marketing Event, It Foreshadows the Future
As Nvidia goes, so goes the entire AI industry. Accelerators, CPUs, networking, security, cooling and power infrastructure, etc, all follow Nvidia’s lead. And at the various GTC events, Nvidia clearly articulates where it is going, and who it is working with to create the future of AI. With these announcements, Nvidia has significantly dialed down the angst around agentic AI security, and accelerated the time to value for its global customer base.
Disclosures: This article expresses the author’s opinions and should not be taken as advice to purchase from or invest in the companies mentioned. Cambrian-AI Research is fortunate to have many, if not most, semiconductor firms as our clients over the years, including Blaize, BrainChip, Cadence Design, Cerebras, D-Matrix, Eliyan, Esperanto, Flex, GML, Groq, IBM, Intel, Micron, NVIDIA, Qualcomm Technologies, Si-Five, SiMa.ai, Synopsys, Ventana Microsystems, Tenstorrent and scores of investment clients. We have no investment positions in any of the companies mentioned in this article except for Nvidia, which we are long.