Korea’s AI semiconductor race is entering a new phase. Dnotitia, a fast-rising deep-tech startup specializing in AI memory and vector data acceleration, is preparing for its IPO after achieving a rare milestone — merging software and semiconductor design into a unified “AI memory infrastructure” model. The company’s path now signals how Korea’s next deep-tech generation may move beyond the limitations of large language models.

Dnotitia Appoints IPO Lead Underwriters as AI Memory Technology Reaches Commercial Stage

Dnotitia has appointed Korea Investment & Securities and Shinhan Investment Corp. as joint lead underwriters for its planned KOSDAQ listing, marking a pivotal step in its transition from research-focused startup to commercialization-ready deep-tech firm.

The company develops Seahorse, an AI agent capable of remembering and retrieving contextual data efficiently, and VDPU (Vector Data Processing Unit) — a proprietary semiconductor that accelerates vector database operations.

Dnotitia completed VDPU’s design and tape-out in December and aims to secure its first silicon chips by June 2026, with full commercialization targeted for early 2027. Despite being a fabless company, it has already generated revenue from its AI software, building a self-sustaining model ahead of hardware production.

AI Memory Infrastructure: Dnotitia’s Core Distinction

Dnotitia’s rise stems from addressing one of the deepest technical limits in modern AI — memory bottlenecks. Traditional LLMs struggle to retain long-term context, leading to inefficiencies in retrieval-augmented generation (RAG) workflows.

The company’s VDPU IP, unveiled at D&R IP & SoC Day, integrates directly into system-on-chip (SoC) structures to enable ultra-efficient vector data processing. According to the company data, its proprietary system can deliver up to 10x faster search speeds and reduce total ownership cost (TCO) by as much as 80% compared with CPU-based software architectures.

By combining hardware acceleration (VDPU) and software orchestration (Seahorse), Dnotitia positions itself as one of Korea’s first AI infrastructure companies offering a full-stack memory solution.

Building Credibility Amid Legal Hurdles and Validation Milestones

Dnotitia’s journey has not been without controversy. In 2025, the company and its CEO Jung Moo-kyung faced legal scrutiny over alleged technology leakage involving former employer Sapeon. Although prosecutors later clarified that that while leaked materials were referenced, they were not found to have been used in developing competing AI semiconductors.

Legal proceedings remain under review, but investors have continued to back the company. Its KRW 100 billion Series A funding under a KRW 300 billion valuation attracted follow-on interest from early investors, reflecting sustained belief in the company’s technical foundation.

In January 2026, Dnotitia was also named among Korea’s Top 100 Emerging AI+X companies by the Korea AI Industry Association, recognized for its semiconductor–software integration. Its Seahorse solution earned TTA’s GS Grade 1 certification, enabling entry into public-sector procurement channels.

Stakeholder Statements: A Market-Ready Transition

Dnotitia CEO Jung Moo-kyung stated,

“Over the past two years since our founding, we have steadily built our technology and business foundation under the vision of becoming a ‘Data System Leader in the AI Era.’
2026 marks the inflection point where our long-prepared technologies and business will begin delivering tangible results. We will continue evolving as a core AI infrastructure company while preparing our IPO in line with this growth trajectory.”

Jung also described Dnotitia’s Mnemos platform, unveiled at CES 2026, as a compact on-device AI agent combining local data use with low-latency reasoning — an early showcase of its memory-first approach.

The company plans to later integrate Mnemos with VDPU and partner NPUs to form an advanced RAG-based agent ecosystem.

Ecosystem Significance: A Different Path from Korea’s Generative AI IPOs

While recent AI IPO attention has centered on generative model developers such as Upstage, Rebellions, and DeepX, Dnotitia represents a different segment of Korea’s deep-tech expansion.

Its focus lies in AI memory acceleration, not model training — an area that forms the infrastructure backbone behind RAG systems and scalable inference.

This distinction makes Dnotitia a critical complement to Korea’s broader AI transformation (AX) agenda, which seeks synergy between semiconductor innovation and enterprise AI deployment.

Industry experts note that companies like Dnotitia could strengthen Korea’s AI sovereignty by securing local IP in vector processing — a field currently dominated by U.S. and Chinese chipmakers.

Its upcoming IPO will thus serve as a litmus test for investor confidence in AI infrastructure startups, bridging Korea’s semiconductor engineering and AI software strengths.

The Dnotitia IPO Preparation: Korea’s Next-Phase AI Infrastructure Signal

If Dnotitia’s IPO proceeds as planned, it would mark one of Korea’s first listings of a company positioned at the intersection of AI memory, semiconductors, and software infrastructure.

The startup’s trajectory — from legal challenges and R&D-intensive chip design to public-market readiness — highlights the emergence of Korea’s second-generation deep-tech ventures, focused not just on algorithms but on the computing systems that sustain them.

For policymakers and investors, the company’s success will test how Korea’s capital markets reward AI infrastructure innovation that extends beyond model-building.

Key Takeaways on Dnotitia IPO Readiness

• Dnotitia, an AI infrastructure startup, has selected Korea Investment & Shinhan Investment Corp. as joint lead IPO underwriters.
• Its proprietary VDPU (Vector Data Processing Unit) enables high-speed vector database processing, addressing AI’s long-term memory limits.
• The company integrates hardware (VDPU) and software (Seahorse) into a unified “memory-centric AI” system.
• Dnotitia completed VDPU’s first design and tape-out, with commercialization targeted for early 2027.
• Despite facing a 2025 legal probe, it retained investor confidence, raising KRW 100 billion under a KRW 300 billion valuation.
• Its Seahorse solution earned TTA GS Grade 1 certification, and Mnemos debuted at CES 2026.
• Dnotitia’s IPO signals Korea’s diversification in AI semiconductors, beyond model developers like Upstage, Rebellions, and DeepX.
• The listing could become a benchmark for Korea’s deep-tech capital readiness and AI infrastructure competitiveness in global markets.

– Stay Ahead in Korea’s Startup Scene –
Get real-time insights, funding updates, and policy shifts shaping Korea’s innovation ecosystem.
➡️ Follow KoreaTechDesk on LinkedIn, X (Twitter), Threads, Bluesky, Telegram, Facebook, and WhatsApp Channel.