Dinotesia said on Jan. 13 it was named in the AI semiconductor category of the Korea AI Industry Association (AIIA)-run "2026 Emerging AI+X Top 100".

"Emerging AI+X Top 100" is a program that selects 100 promising domestic companies expected to lead the future through convergence with a range of industries based on AI technology. The AI semiconductor category selects companies centered on those with specialized semiconductors and data processing technologies to improve AI inference performance.

Dinotesia is developing a data system designed to enable fast and efficient retrieval and use of information and context needed by AI. It addressed data retrieval and memory problems flagged as limitations during the spread of generative AI through a semiconductor-and-software combined structure.

For this, it designed its own vector data operation-dedicated semiconductor, the VDPU (Vector Data Processing Unit). Based on it, the company built the high-performance vector database "Seahorse" and the on-device AI solution "Mnemos". The structure, which integrates software and semiconductors, aims to improve the speed and accuracy of AI services while also improving cost and power efficiency.

A Dinotesia official said AI semiconductors have emerged as a core area of competition along with the spread of generative AI, adding that the selection is meaningful in that Dinotesia's technology direction and industrial potential were officially evaluated.