Yes, a robotics infrastructure supplier could be the next Nvidia, but the path differs fundamentally from semiconductors. While Nvidia built dominance through chip design for AI training, the emerging robotics infrastructure plays—companies enabling the entire ecosystem—are following a similar pattern of consolidation around critical foundational layers. SoftBank’s Roze AI, a data center platform designed specifically to accelerate robotics and AI infrastructure development, exemplifies this shift. The company is targeting a $100 billion valuation with a potential IPO by the second half of 2026, signaling that investors see infrastructure as the bottleneck, not individual robot makers.
The difference is timing and specificity. Where Nvidia could claim neutrality as a chip supplier to competing AI labs, robotics infrastructure suppliers must navigate a more fragmented ecosystem: hardware manufacturers, software platforms, simulation environments, and labor integration. A company that solves the “connective tissue” problem—standardizing how robots integrate data, training, and deployment across industries—would capture disproportionate value, much like Nvidia’s CUDA ecosystem locks in AI developers. This is not speculation: the robotics market is projected to reach $88.3 billion by 2026 and $199.5 billion by 2035, with funding hitting a post-2021 record of $10.3 billion in 2025 alone.
Table of Contents
- Why Robotics Infrastructure, Not Robot Companies, Creates Supplier Moats
- The IPO Boom Signals Infrastructure Consolidation Ahead
- Nvidia’s Physical AI Stack: The Template Robotics Infrastructure Should Follow
- The Supply Chain Leverage Play
- The Geographic Arbitrage and Western Localization Trend
- Funding Velocity Reveals Market Inflection
- The Path to Dominance: What the Next Nvidia Must Do
- Conclusion
Why Robotics Infrastructure, Not Robot Companies, Creates Supplier Moats
The robotics industry is following the classic infrastructure-beats-hardware playbook. Individual robot companies—whether humanoid, industrial, or autonomous—face intense competition and commoditization pressure. Humanoid robotics alone saw over $1.1 billion in funding in 2025-2026, with Apptronik closing a $520 million round in February 2026 and Figure AI raising $675 million at a $2.6 billion valuation. That capital chasing robots drives down margins. Infrastructure suppliers, by contrast, become indispensable to all of them.
nvidia‘s ecosystem is the template. Rather than competing with every AI researcher, Nvidia provides the hardware stack (Jetson modules including the new T4000 with 4x greater energy efficiency), software platforms, and reference implementations that let partners—from Boston Dynamics to Caterpillar to Franka Robotics to LG Electronics—build faster. The NVIDIA Physical AI Data Factory Blueprint is being adopted by companies like FieldAI, Hexagon Robotics, Milestone Systems, and Teradyne Robotics. Once a standard emerges, switching costs rise exponentially. A robotics infrastructure supplier could achieve the same effect by standardizing data pipelines, simulation environments, or robot-to-cloud communication protocols. The bottleneck is rarely the robot itself; it’s the systems that allow robots to learn from shared data and improve across fleets.
The IPO Boom Signals Infrastructure Consolidation Ahead
Two major robotics IPOs in 2026 reveal where capital expects the highest returns. Agibot, a Chinese humanoid robot maker, is targeting a Hong Kong IPO with a valuation between HK$40-50 billion (US$5.1-6.4 billion), expected in Q3 2026 with potential fundraising exceeding $1 billion. The scale is impressive, but Agibot is still a single hardware company. SoftBank’s Roze AI, meanwhile, is positioning itself as the infrastructure layer that makes such companies possible—a data center designed to compress the development cycle for robotics and AI, with a $100 billion target valuation.
this valuation gap is telling. Roze’s higher ambition reflects investors’ belief that infrastructure reaches across multiple industries and geographies, while a single robot company, no matter how capable, is eventually limited to its addressable market. Six new robotics unicorns emerged in March 2026 alone, three of them from China, underscoring the global competition. But unicorns are created, not sustained, without infrastructure. The companies most likely to become multi-trillion-dollar platforms are those that enable all the others—not the ones racing to build the same hardware everyone else is building.
Nvidia’s Physical AI Stack: The Template Robotics Infrastructure Should Follow
Nvidia’s approach to robotics is instructive. The company released new Physical AI models as part of its broader robotics technology stack, specifically designed to work with its hardware and to integrate with partners’ systems. This isn’t just about selling GPUs; it’s about creating a complete layer where robot makers can plug in, reducing their engineering burden and accelerating time-to-market. The Jetson product line—purpose-built edge compute for robotics and autonomous systems—is a superior alternative to generic data center chips because it solves robotics-specific problems: power efficiency, latency, and form factor.
A robotics infrastructure supplier targeting Nvidia-like dominance would need to offer something similarly integrated: not just simulation software, not just data management, but a complete stack that makes it easier for robot manufacturers to build, train, and deploy. The warning: Nvidia itself is already deeply embedded in robotics. Any infrastructure startup aiming for that position must either occupy a different layer (supply chain logistics, fleet management, regulatory compliance) or develop something Nvidia’s current roadmap doesn’t address. Head-to-head competition with Nvidia in robot-specific compute is a losing proposition.
The Supply Chain Leverage Play
Here’s where infrastructure suppliers gain an asymmetric advantage. Approximately 90% of key robotics components are still sourced from China, creating both a bottleneck and an opportunity. Western manufacturers are increasingly motivated to localize production, which requires new infrastructure: local component validation, supply chain aggregation, quality assurance. A company that becomes the hub for localized robotics component sourcing—verifying parts, managing inventory, enabling rapid substitution when supply shocks hit—would wield enormous leverage. This leverage works both ways.
During the semiconductor shortage, companies that controlled the allocation of chips gained more power than the chip makers themselves, temporarily. A robotics infrastructure supplier controlling component access could negotiate better terms with manufacturers while also raising its own valuation multiples. The tradeoff: this requires physical infrastructure (warehouses, testing facilities, logistics networks), not just software. Capital intensity is higher, and regulatory compliance becomes critical, especially if the company handles cross-border supply chains. Roze AI’s focus on data centers is capital-intensive but scalable; a supply chain play would require different economics.
The Geographic Arbitrage and Western Localization Trend
The robotics market’s geographic concentration is shifting. While China continues to dominate component manufacturing and now leads in humanoid robotics investment (three of six new 2026 unicorns), Western companies and governments are actively building alternatives. This creates a temporary opportunity window for infrastructure suppliers. The European and North American robotics ecosystems are consolidating around shared standards and platforms, partly out of supply chain security concerns and partly out of competitive necessity.
A robotics infrastructure supplier with geographic optionality—able to serve North America, Europe, and Asia with localized implementations—could capture value across all three regions. However, geopolitical risk is real. Sanctions, export controls, and data sovereignty rules will increasingly fragment the market. A company offering “Western-friendly” robotics infrastructure could command premium valuations in Europe and North America, but it sacrifices the Chinese market, which is larger and growing faster. Companies must choose: compete globally with compliance complexity, or dominate regionally with pricing power.
Funding Velocity Reveals Market Inflection
The jump to $10.3 billion in robotics funding in 2025—the highest since the 2021 peak—is not driven by hype. It reflects capital redeploying from generative AI (maturing, crowded) into robotics (emerging, less crowded). This funding is concentrating around companies solving systemic problems: humanoid robotics ($1.1 billion), autonomous systems, and increasingly, the infrastructure that ties them together. Figure AI’s $675 million round, with strategic participation from Nvidia and Microsoft, exemplifies the shift toward companies that can build the bridges between hardware, software, and cloud services.
The inflection point is whether this funding sustains. If the robotics hardware market commoditizes (most likely outcome), capital will flee to infrastructure. If specific robot form factors (humanoids, drones, industrial arms) remain differentiated, capital will stay fragmented. Early indicators suggest the former: as humanoid makers proliferate, the competitive advantage shifts from the robot to the software, data, and systems that make robots useful. That’s where infrastructure suppliers win.
The Path to Dominance: What the Next Nvidia Must Do
Becoming the next Nvidia requires solving one critical problem better than anyone else, in a way that becomes harder to undo. For semiconductors, Nvidia solved performance per watt and created CUDA, locking in developers. For robotics infrastructure, the equivalent might be: real-time fleet learning (allowing robots to improve from each other’s data), standardized robot-to-cloud integration (reducing development time by months), or supply chain resilience (guaranteeing component availability when competitors face shortages).
The trajectory is clear: infrastructure suppliers will outpace robot makers in valuation and influence over the next 5-10 years, assuming the industry matures as expected. Roze AI’s targeting of a $100 billion IPO by H2 2026 suggests this shift is already underway. The question for investors and strategists is not whether infrastructure will dominate, but which layers of infrastructure matter most—and which supplier will control them.
Conclusion
The next Nvidia in robotics will not build robots; it will build the systems that make robotics possible at scale. Infrastructure suppliers—data centers, supply chain integrators, software platforms, cloud-to-edge bridges—command higher valuations and longer competitive moats than individual robot manufacturers because they serve the entire ecosystem. With the global robotics market projected to exceed $88 billion in 2026 and climbing toward $200 billion by 2035, the value concentrated in infrastructure is only beginning to materialize.
The race is already underway. Roze AI’s $100 billion valuation target, new robotics unicorns emerging monthly, and record funding levels in 2025 all point to capital understanding this shift. The companies that will define the next decade of robotics are not the ones building the most impressive robots—they’re the ones building the foundation everyone else depends on. For investors, the lesson is simple: watch the infrastructure suppliers, not the robot makers.
