DEEPX, a Seoul-based fabless semiconductor startup, has secured 27 commercial orders across eight countries within seven months of launching mass production of its first-generation AI inference chip. This rapid adoption—uncommon for early-stage semiconductor firms—spans seven Physical AI domains: robotics, smart factories, edge AI servers, industrial AI, surveillance, AI IT services, and smart cities. Deployments now active in Asia, North America, and Europe signal a global appetite for energy-efficient AI hardware.

The company’s supply chain strategy, involving partnerships with Avnet, DigiKey, and WPG, enabled early global distribution. In Europe, Avnet Silica identified 30+ prospective customers for edge AI applications, accelerating regional adoption. Analysts note DEEPX’s ability to scale across multiple countries and industries simultaneously is rare for a fabless startup in its commercialization phase.

Key collaborators include Renesas Electronics, which integrated DEEPX NPUs into industrial boards for smart factories; Sixfab and Raspberry Pi, which developed AI HAT modules; and Ultralytics (YOLO AI model creators), which validated DEEPX NPU compatibility. These alliances position DEEPX’s silicon as an emerging standard for embedded AI solutions.

New products DX-M1M (M.2 module) and DX-AIPlayer (edge acceleration platform) aim to simplify industrial AI deployment. CEO Lokwon Kim emphasized that 350+ proof-of-concept engagements over 12 months ahead of production validated market readiness. With order velocity maintaining pace, DEEPX aims to cement its NPU as the default choice for Physical AI inference—a shift from theoretical concept to tangible reality in embedded systems.