Expedera’s Origin Evolution NPU IP Named Best Edge AI Processor IP in 2026 Edge AI and Vision Product of the Year Awards
April 21, 2026
Highlights
- Expedera’s award-winning Origin Evolution provides memory-efficient, scalable NPU IP for edge-to-data-center AI inference.
- Origin Evolution solves the memory and power bottlenecks that prevent real‑world deployment of GenAI at the edge.
- Expedera’s memory‑efficient NPUs position the company as the go‑to IP partner for SoC teams who want to ship GenAI accelerators without doubling their die size or system power.
SAN JOSE, CA – April 21, 2026 – Expedera, a leading provider of AI accelerator IP for edge and data‑center systems, today announced that its Origin Evolution™ NPU IP has been named “Best Edge AI Processor IP” in the 2026 Edge AI and Vision Product of the Year Awards, presented by the Edge AI and Vision Alliance. The award recognizes innovative building‑block IP powering edge AI and vision solutions, as judged by an independent panel of industry experts and promoted through the Alliance’s platform and the Embedded Vision Summit.
“We designed Origin Evolution to solve the memory and power bottlenecks that prevent real‑world deployment of GenAI at the edge,” said Siyad Ma, CEO and co-founder of Expedera. “Being recognized as Best Edge AI Processor IP by the Edge AI and Vision Alliance validates our approach to right‑sized, memory‑efficient NPUs and reinforces our position as the go‑to IP partner for SoC teams who want to ship GenAI accelerators without doubling their die size or system power.”
Expedera’s Origin Evolution addresses three key challenges for customers developing AI applications on edge devices:
First, faster and more private AI. By running inference on the device rather than in the cloud, you reduce latency and keep sensitive data local, resulting in faster, more secure user experiences.
Second, more performance in the same footprint. Our packet-based architecture breaks networks into packets and routes them through dedicated blocks, reducing external memory transfers by more than 75 percent for popular edge AI models. That lowers system power while boosting throughput.
Third, scalability across your portfolio. You can scale up to 128 TFLOPS per core—and even higher with multiple cores—supporting everything from flagship smartphones and edge devices to cars and data centers, all on one NPU IP platform.
The award trophy will be presented at the 2026 Embedded Vision Summit, May 11-13, 2026, in Santa Clara, CA.
More information about the Origin Evolution NPU IP and its deployment in edge and data‑center AI systems is available at www.expedera.com/origin-evolution-for-edge. Customers and partners interested in licensing Origin Evolution or discussing integration can contact Expedera at [email protected].
About Expedera
Expedera develops high-performance, memory-efficient AI accelerator IP that enables chipmakers to integrate advanced AI and vision capabilities into edge and data center systems. The company’s Origin Evolution NPU IP is designed to run large language models and CNNs with minimal memory bandwidth and power, accelerating the adoption of generative AI in smartphones, automotive, IoT, and data‑center‑class inference platforms.
About the Edge AI and Vision Product of the Year Awards
The Edge AI and Vision Product of the Year Awards recognize leading-edge AI and vision products and technologies, as judged by an independent panel of industry experts convened by the Edge AI and Vision Alliance.
# # #
Media Contact:
Pauline Shulman
+1 650-887-0815
[email protected]
