AI technology by Ceva's NeuPro-Nano gains traction in AI-connected devices market
Ceva, a leading provider of silicon and software IP for smart edge devices, is set to showcase its innovative Ceva-NeuPro-Nano Embedded AI NPUs at CES 2025. The live demonstrations will take place at Ceva's booth, located in the Venetian Ballroom - Bassano 2709.
The Ceva-NeuPro-Nano NPUs are an all-in-one solution for feature extraction, neural network compute, DSP workloads, and control code execution. They are designed for ultra-low power applications, making them ideal for resource-constrained devices in the AIoT and MCU markets.
Key Features and Benefits
The NeuPro-Nano is optimized for energy efficiency, ensuring that devices can run AI tasks without significantly increasing power consumption. It offers high performance for AI processing in a compact form factor, suitable for integration into small devices like edge devices and smart displays. The NPU supports deep learning inference, allowing it to handle complex AI tasks efficiently at the edge.
Thanks to collaborations with companies like Cyberon and AIZIP, there are pre-optimized AI models available for the NeuPro-Nano, enhancing its utility in various AI applications.
Recognition in the AIoT and MCU Markets
Ceva's NeuPro-Nano has gained recognition in the AIoT and MCU markets for its ability to efficiently run AI tasks in low-power devices. This recognition is driven by several factors:
- Ceva has a strong presence in the smart edge technology sector, with over 20 billion devices powered by their IP worldwide.
- Collaborations with companies like ALi and Cyberon have helped integrate Ceva's NPUs into various platforms, enhancing their reach and capabilities in AIoT.
- The integration of Ceva's NPUs into smart edge devices has enabled faster development and deployment of AI-powered solutions, driving innovation in the AIoT space.
Ceva-NeuPro Studio and Edge Impulse Studio Integration
Ceva-NeuPro Studio, an IDE integrated solution for developing, optimizing, deploying, and evaluating AI applications on Ceva-NeuPro-Nano, has been integrated with Edge Impulse Studio. This collaboration enables customers to evaluate AI models on Ceva-NeuPro-Nano before silicon availability, deploy and retrain models for use on Ceva-NeuPro-Nano using NVIDIA TAO Toolkit, streamline development with a unified workflow, and accelerate time-to-market with prototyping and testing.
Customer Wins and Awards
Ceva's latest customer wins are a testament to the success of the Ceva-NeuPro-Nano Embedded AI NPUs in the MCU and AIoT markets. The Ceva-NeuPro-Nano NPUs provide good performance for associated voice, vision, and sensing workloads.
Ceva has won two industry awards: Best IP/Processor of the Year 2024 at the EE Awards Asia event and the 2024 IoT Edge Computing Excellence Award.
Live Demonstrations at CES 2025
At CES 2025, visitors will have the opportunity to see live demonstrations of the Ceva-NeuPro-Nano NPU running embedded AI applications and Edge Impulse Studio. Chad Lucien, the vice president and general manager of the sensor and audio business unit at Ceva, has praised the efficiency of the Ceva-NeuPro-Nano Embedded AI NPUs as an NPU that can simultaneously handle the demands of neural network compute, feature extraction, and complex DSP workloads.
Ceva's commitment to delivering innovative solutions that empower customers to create intelligent, efficient, and scalable edge AI applications is clear in the Ceva-NeuPro-Nano NPUs. Visit Ceva's booth at CES 2025 to experience these groundbreaking technologies firsthand.
[1] Ceva Press Release: Ceva-NeuPro-Nano Embedded AI NPUs [2] Cyberon Press Release: Collaboration with Ceva for AI Solutions
- Visitors at CES 2025 will witness the Ceva-NeuPro-Nano Embedded AI NPUs, integrated with artificial-intelligence capabilities, running embedded AI applications at Ceva's booth.
- Recognized for its energy efficiency and high performance, the Ceva-NeuPro-Nano NPU is designed for resource-constrained devices, making it an ideal choice for artificial-intelligence tasks in the AIoT and MCU markets.
