Rescale at GTC 2026: The Agentic Era Begins with NVIDIA, McLaren and Rescale Leading the Way

The Agentic Era Officially Begins

GTC 2026 was a watershed moment for how AI is already reshaping how work gets done and how new products are developed. Tens of thousands of attendees took part in sessions, exhibits, and demonstrations that revealed a staggering rate of AI adoption in all sectors from industrial engineering to healthcare and life science research. Yes, the hardware and infrastructure became significantly faster and the software got significantly smarter, but it was the use cases and the results that really shined.

Agents, easily the most notable theme at GTC 2026, represent the most practical application of AI the industry has seen yet. McKinsey noted that as the token economy scales and physical AI matures, the organizations that win will be those that govern their data, own their engineering knowledge, and deploy agents that can act on that knowledge continuously. Agents can now break down complex processes, leverage specialized tools, and call on sub-agents to handle tasks in parallel, effectively allowing a single engineer to orchestrate workflows that would previously have required an entire team.

NVIDIA CEO Jensen Huang framed the moment simply: AI is no longer a feature or a tool, it is essential infrastructure. And with that infrastructure now designed from the ground up for autonomous agents, the engineering organizations that fail to adapt risk falling behind companies that are already compressing years of traditional simulation into hours.

The use cases Rescale brought to the event illustrate just how foundational these new technologies can be. We believe this event and the technologies featured are pivotal for the data and AI leaders in engineering looking to bring transformational change to their organizations. 

Rescale and McLaren: Real World Agentic AI and AI Physics Scenarios

Rescale’s partnership with McLaren Automotive was announced at GTC 2026 and showcased through live demos and at the center of the NVIDIA booth that showcased advances in data intelligence, agentic engineering and AI physics. 

McLaren is transforming its product development by embedding true end-to-end agentic AI across its entire engineering lifecycle. Rescale’s digital engineering platform, powered by NVIDIA AI infrastructure, provides an integrated computing, data and AI stack to bring unprecedented speed and scale to McLaren’s design and engineering process.

What does AI look like within a best-in-class engineering organization? The Rescale platform is trained exclusively on McLaren’s own data and integrates NVIDIA AI physics models, agentic engineering libraries, and knowledge bases. This environment connects McLaren’s computer aided engineering (CAE), systems engineering, and design functions into a single, continuously learning platform that adheres to McLaren’s exacting quality standards and performance characteristics, while fundamentally changing how fast the team can move. Agentic workflows on the platform can be generated through conversational prompting, dramatically lowering the barrier to adoption and putting the power of AI-driven engineering directly in the hands of domain experts without requiring deep technical configuration.

Impact Across the Product Development Lifecycle:

• Rapid Design Exploration: Engineers can evaluate thousands of design iterations in hours, covering multiple physics and engineering domains, fundamentally changing how rapidly an optimal performing design can be achieved. Agents broaden the aperture of what is possible, exploring a design space that no human team could cover manually and raising the bar on what engineering excellence looks like.
• Faster Virtual Simulation: AI-driven physics significantly cuts simulation time, with every test feeding new data back into the system to continuously improve surrogate models and agents’ understanding of the physical world.
• Real-Time Performance Prediction: Machine learning models enable near-instant predictions of manufacturing performance, for example in the production of high-performance carbon fibre structures and components.
• Production-Ready Agentic Engineering: Engineering agents handle the highly iterative work that is tedious to configure, troubleshoot, monitor, and interpret individually. They can explore hardware-software optimizations that improve cost efficiency, perform studies scoped to specific budgets, and automate complex, repetitive tasks, boosting expert productivity by 3x on infrastructure workflows and freeing engineers to focus on high-value design thinking.
• Knowledge-Based Engineering: Rescale’s platform builds engineering knowledge graphs that capture insights from previous work, powering agentic workflows and accelerating product development decisions.

Nick Collins, CEO of McLaren Automotive, captured the significance: “This is a genuine strategic transformation for the business. By composing and continuously optimizing our data, our intelligence, and our engineering philosophies at unimaginable speed, we can deliver product developments with agility while protecting the DNA of our company.”

Joris Poort, Founder and CEO of Rescale, added: “Our foundational platform allows McLaren to leverage the latest agentic engineering technologies powered by NVIDIA AI infrastructure, providing a compounding source of competitive advantage for engineers in critical areas such as carbon materials, structural dynamics, and durability, and ultimately the programmatic scaling of engineering excellence across every discipline to deliver world-class products faster.”

The McLaren story is more than an automotive case study. It demonstrates that Rescale’s platform approach to integrating HPC, advanced modeling and simulation and data intelligence is making agentic engineering solutions for simulation, design, testing, and manufacturing available to engineering organizations across industries today.

The Biggest Industry Trends In AI-Driven Engineering

GTC 2026 represented a clear industry pivot: from AI as a productivity layer to AI as an autonomous operating system for engineering and knowledge work.

The Inference Economy is Here

For years, the AI conversation was dominated by training, building larger foundation models on ever-more-powerful clusters. GTC 2026 marked the decisive turn toward inference workloads: the continuous, production-scale deployment of AI put to use for real work in the field. In engineering, this shift is especially significant for AI physics, where inference is what powers surrogate models and physics-based AI to deliver real-time predictions during active design and simulation workflows, rather than as a post-processing step. NVIDIA’s new Vera Rubin platform and Groq language processing unit (LPU) technology are purpose-built for this shift, delivering dramatic improvements in inference throughput and cost efficiency. NVIDIA’s new LPU technology alone delivers up to 35x higher inference throughput per megawatt, a signal that the economics of deploying always-on AI agents are shifting dramatically in favor of adoption.

For engineering leaders, this matters because both the cost and accessibility barriers to deploying AI are falling fast. It is no longer just about giving data scientists and ML engineers better tools. It is about putting powerful AI-driven methods, including AI physics-based inference for modeling, simulation, and design exploration, directly into the hands of engineers who have never written a line of Python. A designer exploring aerodynamic configurations, a materials engineer evaluating composite layups, or a systems engineer validating performance tradeoffs early in the product cycle can now leverage AI physics inference capabilities that were previously the exclusive domain of specialized compute teams. With agents able to perform the iterative, configuration-heavy work in the background and surface results in plain language, the productivity gains extend well beyond the ML team, making every engineer significantly more capable and broadening the range of design possibilities any organization can realistically explore.

The Agentic Ecosystem Is Expanding Rapidly

One of the clearest themes across GTC 2026 was the rapid maturation of the open model ecosystem powering agentic AI. NVIDIA announced the Nemotron Coalition, a growing alliance of AI labs and software platforms contributing specialized open frontier models across language, simulation, robotics, and physical AI domains. This expanding library of open, customizable models is lowering the barrier for enterprises to build purpose-built agents without starting from scratch, while offering better control over model behavior and more favorable economics than closed, proprietary alternatives. Among the most prominent examples is OpenClaw, the open-source agentic framework that has surged in adoption and which Jensen compared to Linux, calling it “the operating system of agentic computers.”

Structured Data as the Foundation for Agentic AI

A consistent theme across GTC 2026 was that structured data is the ground truth of enterprise AI: agents are only as capable as the data they can reliably access and act on. For engineering organizations, this challenge is acute: simulation results, test data, design iterations, and performance logs are often scattered across disconnected tools and teams. Rescale’s data intelligence capabilities address this directly, providing unified data management, metadata enrichment, and lineage tracking across the full engineering workflow, so that agents and AI physics models are operating on clean, contextualized, and governed data from the start. The quality of your AI outputs will only ever be as good as the data foundation underneath them, and for engineering organizations that means investing in three things:

• Unified data fabric that connects simulation, test, and design data 
• Rich metadata and lineage tracking that gives agents the context to act reliably
• Engineering knowledge graphs that capture and compound expertise over time

Managing Change and Maintaining Control Is Critical

As agentic AI matures, governing what agents can access and ensuring they act within sanctioned boundaries has become a central concern. Visibility and control over agent actions cannot be an afterthought and must be considered not just from a technical perspective but embedded in your operational change management. Rescale’s experience guiding agentic engineering deployments ensures that the people and process change complements the enterprise-grade security controls and MCP-based framework that enables granting agents precise permissions, auditing their actions, and scoping their access, so the power of agentic AI can be deployed without putting proprietary engineering data or workflows at risk.

Why These Trends Matter for Engineering Organizations

GTC 2026 drew a clear line in the sand. The shift from training-centric AI to inference- and agent-centric AI is not coming; it is here. The Wall Street Journal noted that inference, the daily use of AI after training, is driving a fundamental shift in how computing infrastructure gets designed and deployed. And the engineering organizations that act on it now will compound advantage over those still evaluating whether to start.

Rescale has been building toward this moment. Our platform was designed for engineering organizations that need to move fast without sacrificing the rigor that defines world-class products. The McLaren partnership is the clearest expression of that mission: a company with 60 years of engineering heritage using Rescale to compress product development timescales, scale their best engineers’ instincts programmatically, and protect the institutional knowledge that defines their brand.

Learn More About Rescale’s GTC 2026 Announcements

Ready to see what Rescale’s digital engineering platform can do for your organization? Visit our GTC 2026 page for the McLaren announcement, Rescale demo resources, and information on how to get started. Explore Rescale’s GTC Resources and see Rescale featured in NVIDIA’s GTC recap blog.