By Ian Werther in News — 13 Mar 2026

KRIBB, Puriosa AI Sign MOU for NPU-Based AI Drug Discovery

Seoul, South Korea — the Korea Research Institute of Bioscience and Biotechnology (KRIBB) announced a push to accelerate AI-driven drug discovery, signing a memorandum of understanding with Puriosa AI to establish a research collaboration focused on neural processing unit (NPU)–based AI drug development.

The core of the agreement is to actively apply Puriosa AI’s second-generation NPU, RNGD, to digital AI cells (virtual cells) and a digital bio platform, expanding the computational tools available for AI-assisted drug design and screening.

The Regulatory Fish Encyclopedia (RFE) project (1993-2000) was a compilation of data in several formats to assist with the accurate identification of fish species. It was developed by FDA's RFE Team of scientists at the (former) Seafood Products Research Center, Seattle District, and the (former) Center for Food Safety and Applied Nutrition to help federal, state, and local officials and purchasers of seafood identify species substitution and economic deception in the marketplace. The RFE included high-resolution photographs for a number of commercially relevant fish species for sale in the U.S. market. The scanned digital images include whole fish as well as their marketed product forms such as fillets, steaks, etc.
Images used in the RFE were photographed on a modified Polaroid MP4 camera system fitted with a Polaroid 545 holder for 4" x 5" film, and a 135 mm lens. Quickload tungsten-balanced ISO 64 E-6 process 4" x 5" transparency film was used to capture the images. A 10 cm scale bar was included in each photograph to indicate specimen size and magnification, and a standard Kodak color bar was included to assure color accuracy. Developed transparencies were professionally scanned to produce the digitized graphic images provided here. — Representative image for context; not directly related to the specific event in this article. License: Public domain. Source: Wikimedia Commons.

The collaboration also brings in Yonsei University College of Pharmacy, Omphalos Korea, and Insilicox, alongside KRIBB’s National Center for Preclinical Trials (KPEC) and other partners. The effort is organized under the “Digital AI Cell Sejong the Great Project Consortium,” which is led by KRIBB’s Director Ko Kyung-chul, with research leadership from KRIBB’s center.

Over time, the partners aim to strengthen a research network spanning industry, academia, and government to advance AI-enabled drug discovery. Planned areas include AI-based drug development, upgrading digital AI cells and the digital bio platform, development of biotech-focused large language models and data analytics platforms, and the creation of next-generation biotech logic tailored to lab environments. They also seek to integrate data analysis platforms with computing farm resources.

KRIBB’s Ko Kyung-chul underscored that AI drug development relies on large-scale bioinformatics—data analysis, AI model training, and virtual screening—and that introducing high-performance, NPU-based AI computing infrastructure will bolster digital-bio research capabilities, including identifying drug candidates and predicting drug responses.

Structural biology plays a key role in drug discovery. First, when developing a new drug treatment for a disease, gene targets can be found using genomics. Second, those target proteins can be analyzed to find appropriate sites for the new drug to bind. Finally, a new drug can be developed that matches those specifications based on the chemical library. — Representative image for context; not directly related to the specific event in this article. License: CC BY-SA 4.0. Source: Wikimedia Commons.

Puriosa AI CEO Jun-ho Baek said the convergence of AI semiconductors and biotech is rapidly altering drug development, and the collaboration aims to contribute to building the next generation of AI computing infrastructure needed for a digital-bio era.

Beyond Korea, the partnership signals an expanding trend in AI-enabled life sciences that matters to U.S. readers for several reasons. It highlights growing cross-border collaboration on AI drug discovery, a key area for accelerating pharmaceutical innovation and potentially shortening development timelines. The effort also reflects broader demand for advanced AI hardware and specialized software platforms that could influence global supply chains for biotech and pharmaceutical products, including any future joint ventures, licensing, or investment flows involving U.S. partners. Finally, the push to develop biotech-specific AI models and data platforms may shape international standards and interoperability in bioinformatics, data sharing, and regulatory frameworks that affect U.S. research institutions, startups, and large pharmaceutical companies.

Subscribe to Journal of Korea