Expedera

Expedera develops neural network processor (NPU) IP cores that provide AI inference acceleration for edge devices, mobile, automotive, and data center applications. The company's Origin platform uses a unique packet-based architecture that subdivides each neural network layer into self-contained, independently executable fragments, enabling parallel execution across multiple layers, better resource utilization, and deterministic performance with up to 90% processor utilization. This approach reduces memory requirements to bare minimum and eliminates the need for hardware-specific optimizations, allowing customers to run trained neural networks unchanged without reducing model accuracy. The product portfolio includes Origin E1 (1 TOPS for always-on applications), E2 (1-20 TOPS for mobile and edge), E6 (16-32 TOPS for smartphones and AR/VR), E8 (32-128 TOPS for automotive and data center), and TimbreAI T3 (ultra-low power for noise reduction in headsets). The Origin Evolution line supports popular LLM and CNN networks across automotive (scaling to 96 TFLOPS), data center (128 TFLOPS), edge (32 TFLOPS), and mobile (64 TFLOPS) configurations, with multi-core scaling to PetaFLOPS. The technology achieves 18 TOPS/W efficiency and is field-proven in over 10 million consumer devices. Founded in 2018 and headquartered in Santa Clara, California, Expedera has raised $38 million in a Series B round backed by Indie Semiconductor and Ignite Innovation.

Edge Computing > Edge AI Processors > AI Accelerators,

AI Infrastructure > Enabling Technologies > Hardware > Processors > AI Accelerators,

Photonics > Optoelectronics > Processors > AI Based,

Embedded Systems > Embedded Hardware > Core Processors > Application Specific Processors > AI Accelerator,

Digital ICs > Microprocessors > Artificial Intelligence and Machine Learning > Optical Processing Unit > Neuromorphic

Company Wise > Cisco,

College Wise > IIT Kanpur