As AI workloads continue shifting toward intelligent edge devices, software platform companies are increasingly investing in custom silicon to optimize performance, power efficiency, and user experience. Ceva has announced a landmark AI licensing agreement with a major U.S. software and AI platform company to power a custom AI silicon program, marking a significant expansion of its customer base beyond traditional semiconductor manufacturers and device OEMs.
Ceva has entered into a strategic AI licensing agreement with a major U.S. software and AI platform company for the development of custom AI silicon targeting next-generation intelligent computing devices. The agreement represents an important milestone for the company, extending its licensing business beyond traditional chipmakers and original equipment manufacturers as software platform providers increasingly design proprietary silicon to enhance AI performance and efficiency.
The customer selected Ceva's NeuPro-M Neural Processing Unit (NPU) IP as the foundation for its custom AI silicon architecture. NeuPro-M is designed to deliver scalable, power-efficient AI acceleration for advanced on-device inference, supporting generative AI, multimodal AI, emerging agentic AI applications, and other machine learning workloads while operating within the strict power, thermal, and area constraints of intelligent edge devices.
According to Ceva, the collaboration includes joint optimization of neural network implementations tailored to the customer's target workloads. By enabling tighter integration between silicon, operating systems, AI software frameworks, and hardware, the solution aims to improve performance, reduce power consumption, and enhance the overall user experience compared with off-the-shelf processors.
"The decision by one of the industry's leading software and AI platform companies to build custom AI silicon on NeuPro-M reflects a broader shift toward AI-first computing architectures."
"We view this as one of the most strategically significant AI licensing agreements in Ceva's history, reflecting the growing role of AI acceleration in shaping the future of computing."
The agreement highlights a broader industry trend in which software platform companies are investing in proprietary AI chips to achieve greater control over performance, power efficiency, and hardware-software integration. As AI workloads continue expanding across edge devices, custom silicon is becoming an increasingly important differentiator for delivering advanced AI capabilities in portable and intelligent computing systems.
The landmark licensing agreement reinforces Ceva's position in the intelligent edge AI ecosystem while demonstrating the growing importance of specialized AI processors for next-generation computing platforms. By extending its technology into custom AI silicon programs for software platform companies, Ceva is positioning its NeuPro-M architecture to support the next wave of AI-enabled devices and edge intelligence.
About Ceva
Ceva is the leading licensor of silicon and software IP that enables Smart Edge devices to connect, sense, and infer data more reliably and efficiently. The company's portfolio includes wireless communications, sensing, and Edge AI technologies that power semiconductors for consumer, industrial, automotive, mobile, PC, and infrastructure applications. Through its scalable AI processors and software platforms, Ceva enables customers to develop intelligent devices optimized for performance, power efficiency, and advanced AI capabilities.
TechIntelPro empowers B2B technology decision-makers with the latest industry trends, deep insights, leadership perspectives and actionable intelligence to drive both organizational and professional milestones, and thrive on the tech pulse.
