Ethernovia Secures $90M in Series B Funding to Develop AI Networking Hardware

Ethernovia has secured over $90 million in a Series B funding round, fueled by growing demand for networking semiconductors that power real-time autonomy in vehicles, robots, and industrial systems. The round was led by Maverick Silicon, with investments from Socratic Partners, Conduit Capital, and CDIB-TEN Capital. Existing backers such as Porsche SE, Qualcomm Ventures, and Fall Line Capital also participated.

Headquartered in Silicon Valley, Ethernovia is creating a new category of Ethernet-based packet processors to serve as the central data backbone—or "nervous system"—for intelligent edge machines. The company addresses a critical bottleneck in autonomy: transporting massive volumes of sensor, vision, and AI data reliably and efficiently in real time.

Redesigning the Data Backbone for Autonomous Systems

Autonomous vehicles, advanced driver-assistance systems, and industrial robots increasingly depend on dozens of high-bandwidth sensors and AI calculations that require precise, predictable timing. Legacy in-vehicle and industrial networks were not designed for these demands, often leading to fragmented architectures, greater system complexity, and rising costs.

Ethernovia focuses on packet processor-driven, Ethernet-based architectures that unify networking, computation, and data management. Instead of relying on a collection of outdated communication buses and point-to-point connections, its platform is built to aggregate and route real-time data streams in a programmable, scalable manner—supporting both zonal and centralized system designs.

Packet Processors Engineered for Physical AI

The core of Ethernovia's technology is a family of high-performance packet processors built specifically for edge and physical AI applications. These chips are engineered to handle high-bandwidth sensor and AI traffic with deterministic latency and high power efficiency—two factors increasingly critical for success in automotive and robotics.

By enabling programmable data paths and scalable Ethernet fabrics, the platform supports software-defined systems that can evolve via over-the-air updates while maintaining the performance required for safety-critical operations. This adaptability is especially important as automakers adopt architectures where functionality is determined more by software than by fixed hardware.

Growth Across Automotive, Robotics, and Industrial Markets

While automotive is a primary market, Ethernovia's technology targets multiple sectors where real-time edge intelligence is essential. Robotics, industrial automation, and new AI-driven machines all face similar challenges with latency, synchronization, and data movement. In these fields, performance is often limited not by raw computing power, but by how effectively data can travel between sensors, processors, and actuators under strict timing constraints.

These industries are also undergoing architectural convergence. Robotics and industrial systems are beginning to adopt principles once unique to automotive, such as zonal architectures and centralized computing, while automotive platforms integrate data center concepts like software-defined networking and standardized Ethernet fabrics. This trend is driving demand for networking semiconductors that perform reliably in diverse environments, support long product lifecycles, and accommodate evolving software needs.

The new capital will accelerate development and production of Ethernovia's next-generation packet processors, expand its software and system capabilities, and deepen customer engagements across these industries. As projects move from pilots to full-scale production, the focus is shifting toward platforms that support long-term upgrades, mixed workloads, and increasing autonomy without needing complete redesigns.

What This Means for the Future of Physical AI

Ethernovia's successful funding round underscores a broader shift in autonomy and robotics: intelligence is no longer limited just by algorithms, but by the infrastructure that connects sensing, decision-making, and action in the physical world. As AI moves from the cloud into vehicles, factories, and machinery, networking semiconductors are becoming a foundational layer, not an afterthought.

This change reflects a growing awareness that physical AI systems are, at their core, real-time systems. Delays, lost data packets, or unpredictable latency can lead to tangible problems, from reduced performance to safety hazards. Consequently, reliable, predictable data movement is becoming as important as model accuracy or computing power.

Packet-centric, Ethernet-based architectures suggest a future where intelligent machines are more modular, upgradable, and software-defined, similar to the evolution seen in data centers over the last ten years. If this transition continues, competition in physical AI may increasingly depend on which companies can provide the most reliable and adaptable data fabric—one that supports continuous innovation without compromising real-time performance.