Following LG, Naver also 'meets NVIDIA'... K-AI's presence grows

Following LG, Naver also 'meets NVIDIA'... K-AI's presence grows

NVIDIA VP Catanzaro visits Naver 1784 office

NVIDIA unveils Korea-specific Nemotron dataset, expands ecosystem

NVIDIA is actively strengthening its technology alliance with major domestic AI companies such as Naver and LG to expand its open artificial intelligence (AI) ecosystem.

According to the IT industry on the 23rd, Brian Catanzaro, NVIDIA's VP of Applied Research, who recently visited Korea, met with Sung Nak-ho, Head of Naver Cloud Hyperscale AI, at Naver's 1784 office in Seongnam, Gyeonggi Province, the previous day.

Both sides focused on discussing cooperation measures for the domestic expansion of NVIDIA's next-generation open AI ecosystem 'Nemotron' and the advancement of Naver Cloud's 'HyperCLOVA X' follow-up model.

At the meeting, VP Catanzaro reportedly highlighted the partnership and technological synergy for Nemotron ecosystem expansion, paying close attention to Naver Cloud's Korean language data and optimization technology.

Naver Cloud also proactively proposed improvements to enhance technological completeness, based on a low-precision computing technique that increases learning speed by over 10% compared to existing methods using its own pipeline.

From the 21st, NVIDIA has been holding 'Nemotron Developer Days' in Korea for four days, embarking on expanding the Nemotron ecosystem, which encompasses open-source based AI models, datasets, and software.

In particular, NVIDIA plans to support developers in building AI systems that better understand Korean contexts by releasing the Korea-specific synthetic dataset 'Nemotron Persona Korea' on Hugging Face for commercial use.

This dataset was built based on public data from sources such as the National Statistical Portal (KOSIS), the Supreme Court, the National Health Insurance Service, and the Korea Rural Economic Institute. Naver Cloud also provided foundational data and domain expertise from the design stage.

The industry believes that NVIDIA has fully embarked on a strategy to move beyond simple model performance competition, encouraging each country's AI ecosystem to be built on its platform using country-specific datasets.

Domestic companies are also expanding their scope of cooperation in line with this.

LG AI Research agreed on April 21st to jointly develop specialized models incorporating the Nemotron open ecosystem during a meeting with VP Catanzaro, and SK Telecom [017670] is also actively utilizing NVIDIA datasets and solutions in the development of its next-generation model 'A.X K2'.

In addition, major domestic AI model and infrastructure startups such as Upstage, Elice Group, and Motif Technologies also participated in NVIDIA's event, expressing their willingness to expand partnerships.

A Naver Cloud official stated, "In this collaboration, we supported the accurate reflection of Korean characteristics and cultural contexts in AI models, not merely as users but as co-developers," adding, "We will accelerate the advancement of our AI models based on this cooperation."