Mark Goranson is the Chief Executive Officer EMASS, a wholly owned subsidiary of Nanoveu, for which they serve as the semiconductor technology division. With more than 45 years in the global semiconductor industry, he has held senior leadership roles at companies including Intel, Freescale Semiconductor, and ON Semiconductor. He specialises in wafer fabrication, assembly, testing, and large-scale manufacturing and is responsible for driving EMASS’s commercialisation of ultra-low-power edge-AI system-on-chip solutions. He joined following EMASS’s acquisition by Nanoveu, where his expertise in scaling and strategic alliances is helping accelerate time-to-market for the company’s next-generation SoC offerings.
Tell us about your company?
EMASS is building the next generation of ultra-low-power Edge AI semiconductors. Our mission is to enable intelligent devices to continuously sense, understand, and respond to their environments without relying on the cloud and without sacrificing battery life. Through our ECS-DoT platform, we bring real-time AI directly onto wearables, drones, industrial systems, smart infrastructure, and other battery-powered devices. We believe the future of AI is not only in data centers—it is at the edge, where decisions need to be made instantly, privately, and efficiently. This shift toward always-on intelligence is driving the next wave of innovation, and EMASS is focused on enabling that transition.
What problems are you solving?
Battery-powered edge devices must continuously input sensor data and respond in real time within extremely limited power budgets. Many conventional edge processors process most if not all inputs, whether meaningful or not. They also rely on duty-cycled operation, external memory, or cloud connectivity to handle AI workloads, which can increase latency, consume more power, create security and privacy concerns, and reduce battery life.
EMASS developed ECS-DoT to enable always-on, real-time AI inference directly on-device while maintaining ultra-low power consumption and low latency. Its event-driven system-on-chip architecture keeps devices in an ultra-low-power state until meaningful sensor activity is detected, reducing unnecessary computation and energy use. By processing AI models and sensor data entirely on-chip, ECS-DoT removes the need for continuous cloud-based inference or frequent external memory access. This enables reliable, real-time decision-making for applications like voice detection, motion sensing, condition monitoring, and autonomous response, while extending battery life and reducing system complexity.
What application areas are your strongest?
EMASS’s ECS-DoT delivers the greatest value in battery-powered systems requiring continuous sensing and real-time intelligence within strict energy constraints. In wearables like smart glasses, hearables, and fitness trackers, ECS-DoT enables ultra-low-power, always-on processing for private voice interfaces, gesture detection, and biometric monitoring. One use case uses bone-conduction sensing through an IMU to support keyword spotting without continuously active microphones, improving power efficiency and privacy. Battery-powered drones and robotics also benefit from real-time onboard AI inference within a sub-milliwatt power envelope. In validation trials, this ultra-efficient processing improved operational endurance, making the technology well suited for industrial inspection, agriculture, surveying, and public safety. In industrial IoT and smart infrastructure, ECS-DoT processes sensor data locally to detect anomalies, monitor equipment conditions, and support event-driven wireless communication. Transmitting only meaningful events reduces bandwidth requirements, extends battery life, and enables scalable monitoring of remote infrastructure.
What keeps your customers up at night?
Battery life. For someone building a medical wearable, battery life can be critical – that battery should last as long as possible before it has to be replaced. Or maybe you can design with a smaller, lighter, less obtrusive battery for the next generation of product and still get the same amount of battery life. For someone building autonomous or semi-autonomous systems, such as drones, battery life might be less an issue of replacing the battery and more about utilization rates – the less time a unit spends recharging is that much more time it’s being used for its purpose. And here’s a brand new consideration: until now, AI has been used in power-constrained devices to evaluate what the device’s sensors are detecting. We can make that more efficient, but with our system we’re also making it practical to use AI for the operation of the device itself. Using a drone as an example, you can use AI to significantly improve flight dynamics. What keeps everyone up at night? “How can I compete better?” We can help in more than one way.
What does the competitive landscape look like and how do you differentiate?
The edge AI market is crowded with solutions designed for smartphones, cameras, and devices with relatively large batteries and power budgets. Many processors can run AI, but very few can do so continuously within a sub-milliwatt power envelope.
EMASS was designed specifically for always-on intelligence. ECS-DoT combines event-driven sensing, integrated memory, dedicated AI acceleration, and ultra-low-power operation to deliver real-time inference while consuming a fraction of the energy required by conventional approaches. In many applications, this allows customers to deploy AI capabilities that would otherwise be impractical because of battery life, thermal constraints, size limitations, or privacy concerns.
Our focus is not simply making AI faster—it is making AI continuously available in devices where every milliwatt matters. That is a fundamentally different design objective than most competing solutions.
What new features/technology are you working on?
Three tracks. On silicon, we’re scaling power and performance on more advanced process nodes and extending the always-on architecture across richer multi-modal workloads — vision, audio, and motion together. On software, we’re putting real investment into the developer experience: an expanding model zoo of ready-to-deploy models, a toolkit and IDE that shorten the path from a customer’s data to a model running on the chip, and broader framework support so teams can bring their own models easily. Third, we’re deepening partnerships across sensor makers, distributors, and toolchain partners, so ECS-DoT drops cleanly into real designs. The common thread is making always-on edge AI not just possible, but genuinely easy to adopt.
How do customers normally engage with your company?
Almost always through an evaluation first. We get a developer kit into the customer’s hands so they can run ECS-DoT in their own environment, benchmark the power against their target, and start integrating with our SDK. From there, the serious opportunities move into a design-in phase — a proof-of-concept or reference design built around their use case, then integration, design review, and commercial terms. We support that with reference designs across our core verticals and with field engineers plus a global network of distributors and reps. We also work shoulder-to-shoulder with sensor partners and device makers during integration. My goal is that the chip shows up as part of a working system — not a part the customer has to go figure out on their own.
Also Read:
Share this post via:
RISC-V and AI: The Architecture Shift Is Now