NXP Semiconductors has agreed to acquire US edge AI chipmaker Kinara for $307 million. Kinara specializes in energy-efficient neural processing units (NPUs). NXP said it will more formally integrate the company's edge NPUs and AI software into its own industrial and IoT processors; the two companies already work together. They are pitching IoT and AI systems for the industrial and automotive markets. The deal is expected to close in the first half of this year.

Kinara's NPUs, including the Ara-1 and Ara-2, are among the best in class for performance and energy efficiency. They are used for emerging AI applications in vision, voice, gesture, and a variety of other generative AI systems. Both support mapping inference graphs on their proprietary neural processing units. Their programmability ensures adaptability as AI algorithms continue to evolve from CNNs to new approaches such as generative AI and agent-based AI in the future.

Its NPUs and software deliver energy-efficient AI performance across a range of neural networks, both traditional and generative AI. This includes "tiny machine learning" (tinyML) systems that are miniaturized to run even on low-power IoT sensors. NXP said the deal will enhance its ability to offer a "complete and scalable AI platform" by bringing "standalone NPUs and powerful AI software"to its portfolio of processors and other solutions.

NXP acquires edge AI pioneer Kinara to redefine the intelligent edge

NXP Semiconductors N.V. (NASDAQ: NXPI) today announced that it has entered into a definitive agreement to acquire Kinara, Inc., an industry leader in high-performance, energy-efficient and programmable standalone neural processing units (NPUs). These devices support a broad range of edge AI applications, including multi-modal generative AI models. The acquisition, which will be made in an all-cash transaction valued at $307 million, is expected to close in the first half of 2025, subject to customary closing conditions, including regulatory approvals.

The future of intelligent systems will require secure, cost-effective and energy-efficient edge AI processing. As a result, the edge AI processing market is growing rapidly. Advanced edge AI enables critical decisions to be made locally, independent of the cloud, resulting in faster response times, improved data privacy and lower costs.

Kinara's innovative NPUs and comprehensive software support deliver energy-efficient AI performance for a range of neural networks, both traditional and generative AI, to meet the rapidly growing AI needs of the industrial and automotive markets. The acquisition will enhance and strengthen NXP's ability to offer a complete and scalable AI platform, from TinyML to Generative AI, bringing standalone NPUs and powerful AI software to NXP's portfolio of processors, connectivity, security and advanced analog solutions.

As existing partners, Kinara and NXP can easily pair Kinara's NPUs with NXP's industry-leading portfolio of industrial and IoT processors. Together, the two companies will create tighter solution integration to deliver a scalable AI platform for a wide range of industrial and automotive AI inference needs.

"The industrial market is undergoing a transformation, with innovative technologies such as Generative AI helping to significantly improve efficiency, sustainability, safety and predictability, and in many cases, unlocking new use cases and capabilities," said Rafael Sotomayor, executive vice president and general manager of Secure Connected Edge at NXP. "Adding Kinara's AI capabilities to our broad portfolio of intelligent edge products creates a scalable platform for new AI systems. Together, we can help customers simplify complexity and accelerate time to market as they create transformative AI systems."

About Kinara

Kinara's standalone NPUs, including Ara-1 and Ara-2, lead the industry in performance and power efficiency. This makes them the solution of choice for emerging AI applications in vision, voice, gesture, and a variety of other generative AI-driven multimodal implementations. Both devices feature an innovative architecture that maps inference graphs for efficient execution on Kinara's programmable proprietary neural processing units, maximizing edge AI performance. This programmability ensures adaptability as AI algorithms continue to evolve from CNNs to new approaches such as generative AI and agent-based AI in the future.

Ara-1 is the first generation of standalone NPUs capable of advanced AI inference at the edge. Ara-2, the second generation NPU, is capable of up to 40 TOPS (tera operations per second) and is optimized for system-level high performance for generative AI. Ara-1 and Ara-2 NPUs can be easily integrated with embedded systems to enhance their AI capabilities, including upgrading existing field systems.

Kinara also offers a complete software development kit that enables customers to optimize AI model performance and simplify deployment. Kinara's AI software portfolio includes an extensive model library and model optimization tools that will be integrated into NXP's eIQ AI/ML software development environment, enabling customers to quickly and easily create end-to-end AI systems.

Source: Content from nxp and others

View more at EASELINK