​CEVA Unveils 2nd-Gen NeuPro NPU, Doubling AI Throughput

What happened

CEVA unveiled its second-generation NeuPro neural processing unit (NPU) product lineup on Nov. 2. Compared with the first generation, the second-generation lineup delivers up to twice the AI processing throughput per square millimeter, with energy efficiency improved by 50%. The lineup is currently available to select customers for early access, with full production supply scheduled to begin in the second quarter of 2027. CEVA plans to expand its business into AI infrastructure, including AI inference semiconductors, intelligent memory, and connectivity chips, using the new product family. More than 15 customers have licensed the first-generation NeuPro. All products in the lineup use the same instruction set architecture (ISA) and a common software development kit (SDK), support applications from low-power sensors and edge devices to data centers, cover workloads from object recognition to transformer-based LLMs, multimodal AI, agentic AI and Physical AI, and support FP4 and FP8 floating-point formats along with mixed-precision computing.

Impact and signal

Affected components and suppliers

Source and time

Source: The Elec · 11/2/2026 · View original source