Jason Hardy
Chief Technology Officer for Artificial Intelligence
August 27, 2025
The robots we imagined in science fiction are no longer fiction. They’re optimizing power grids, managing transportation networks and supply chains, and transforming manufacturing floors. The question isn’t whether AI will reshape these and other industries. Rather, it’s how quickly organizations can bridge the gap between digital intelligence and physical reality to be among those leading the charge.
The AI Evolution: Four Waves of Intelligence
AI has rapidly evolved through a series of distinct phases defined in different ways. I happen to like NVIDIA’s overview shared in the CES keynote this year. It begins with perception AI, which taught machines to see (e.g., analyzing medical images, recognizing patterns and speech, and processing video feeds). Next came generative AI, which gave machines creativity — the ability to synthesize inputs and then use it to create content and expand concepts from simple prompts.
Then most recently (this is where things get really fun) came agentic AI, which enables intelligent agent-driven planning, decision making, workflow automation and process optimization.
Now comes the fourth phase, physical AI, with intelligence that doesn’t just perceive or create, but acts. These systems control autonomous vehicles and robots, optimize smart spaces and manufacturing processes, and manage critical infrastructure, all in real-time.
Accelerating Physical AI With NVIDIA RTX PRO Servers
Of the AI phases, it’s this physical AI wave that has the greatest potential to change how the world works. That’s because it connects all the operational and informational threads across a broad range of major global industries, such as energy, transportation, and manufacturing, all areas the Hitachi Group focuses on.
From that perspective, the convergence of digital intelligence with physical systems changes everything, enabling AI solutions that can understand, predict, and control physical processes with speed and precision that surpasses human capability.
Hitachi is among the first companies in any industry to deploy NVIDIA RTX PRO Servers, built on its NVIDIA RTX PRO 6000 Blackwell Server Edition GPU. The servers are designed to accelerate the most demanding agentic and physical AI workloads. This allows organizations to transform their data center infrastructure to support AI-powered industrial applications.
In NVIDIA’s announcement, Hitachi, Ltd. CEO Toshiaki Tokunaga summed up the significance to our organization and customers, noting that with NVIDIA RTX PRO, “Hitachi will enable the digital twin and optimization of physical assets, including social infrastructure through the acceleration of AI reasoning and physical AI, while also unlocking new possibilities such as improving productivity across overall business activities.”
Digital Twins: Where Simulation Meets Reality
Digital twins are foundational to this transformation — virtual replicas of physical systems that enable real-time simulation, optimization, and predictive maintenance. But creating them requires computational power that must handle complex physics, multiple variables, real-time interactions, and photo-realistic renderings simultaneously — adding layers of complexity that traditional systems struggle to manage.
The IT infrastructure challenge is real. Recent advances in GPU technology — NVIDIA RTX PRO 6000 Blackwell GPUs being a prime example — are finally addressing the computational demands of the most complex industrial AI application. And not a moment too soon.
Three critical areas showcase this potential:
Platforms like NVIDIA Omniverse are enabling the creation of photorealistic digital twins and collaborative 3D workflows. Previously, these projects often faced significant limitations due to long runtimes and limited GPU availability in cloud environments.
Modern GPU infrastructure allows for rapid iteration on large 3D models, creating digital twins that are not just visually accurate but physically precise. These simulations can model everything from robotic movements to power grid fluctuations with unprecedented detail and speed.
For Hitachi Vantara, the implications are significant. To develop effective physical AI systems, enterprises need to test and optimize their robotic fleets in simulation before deploying to real-world factories. Working with NVIDIA RTX PRO Servers and NVIDIA Blackwell GPU technology, and solutions like Hitachi iQ, we offer our customers advanced industrial and physical AI solutions, delivering faster performance for digital twin, simulation, and synthetic data generation workflows. Combined with our deep operational technology (OT) expertise, we’ll be able to help them unlock breakthrough solutions in Omniverse-based digital twins and physical AI.
Data Center Performance and Control. Cloud Scalability and Agility.
Organizations developing industrial AI face a critical decision — the pull of cloud-based/OpEx agility or on-premises/CapEx control.
The math is compelling — for AI development workloads, on-prem data center infrastructure can deliver 40-60% cost reductions compared to cloud spending.
But cost isn’t the only factor weighing in the data center’s favor. Mission-critical AI solutions for power grids or transportation systems require guaranteed access to high-performance computational resources. The unpredictability of the cloud doesn’t align with industrial reliability requirements.
Additional challenges of cloud-based AI solutions compound the problem:
When you add it all up, data centers should easily win. But they haven't always. Traditional on-premises solutions were often oversized and overpriced for initial AI exploration — creating a barrier to adoption.
The modular nature of NVIDIA RTX PRO Servers changes this dynamic. It enables us to work with a variety of customers across numerous use cases, like helping large enterprises to start small in a controlled environment, then go big once the system is ready to scale.
“We’ll leverage the RTX PRO Server with Hitachi iQ to meet evolving customer demands, especially in addressing complex industrial and enterprise AI challenges," Octavian Tanase, Hitachi Vantara Chief Product Officer, says. "This enables faster iteration, richer simulations, and the creation of AI solutions that are production-ready from day one.”
The AI Industrial Revolution Starts Now
Recently, NVIDIA's President and CEO Jensen Huang said that "enterprises everywhere are re-architecting their data centers for the AI industrial revolution." The transition from general-purpose computing to accelerated computing infrastructure represents a fundamental shift in how enterprises need to approach technology for the AI era.
This transformation extends beyond any single organization. And it isn’t just about adopting new technology. It’s about positioning organizations at the intersection of digital intelligence and physical reality. As AI systems become more capable of understanding and manipulating the physical world, companies that combine deep operational technology expertise with cutting-edge AI capabilities will likely define the future of industrial automation. Physical AI transforms automation from a collection of smart machines into a unified nervous system for civilization itself. When every factory, power plant, and transportation hub can communicate, coordinate, and collectively optimize in real-time, we’re not just changing how things are made — we’re fundamentally altering how society functions at scale.
The science fiction vision of intelligent robots and autonomous systems is rapidly becoming reality. But rather than replacing human workers, these systems are augmenting human capabilities, making industrial processes safer, more efficient, and more sustainable.
As we look toward a future where AI doesn’t just process information but actively shapes our physical world, investments in physical AI infrastructure will determine which organizations lead this transformation rather than simply respond to it.
The robots may not look exactly like those in “Blade Runner” or “Battlestar Galactica” but their impact on our daily lives is becoming just as profound. The question facing every company, in every industry, is how to ensure this AI-powered future serves humanity’s most critical needs. Including reliable power, safe transportation, and efficient manufacturing.
As we all work towards developing and deploying innovative solutions across the entire AI spectrum, it’s essential to have immediate access to cutting-edge software and infrastructure to ensure the best possible enterprise AI outcomes. The right partners are equally important.
Reach out if you’d like to discuss. Always happy to be in good company.
Read More:
Jason Hardy is Chief Technology Officer for Artificial Intelligence at Hitachi Vantara. Connect with Jason on LinkedIn.
