Internet

From GPU Center to ‘Smart Factory’… Naver “Accelerates Asia-Pacific AI Factory”

Kim Ji-hoon, Director at Naver Cloud “Combining the Strengths of NeoCloud and Hyperscalers” “Demand for AIDC Expected to Rise as Inference and Physical AI Gain Traction” “Preparing Key Hubs for AI Factories in the Asia-Pacific Region”

Shin Yeong-bin
2026-10-02 08:55:02
[Edaily Reporter Shin Yeong-bin ] Naver (NAVER(035420)) is accelerating its “Artificial Intelligence (AI) Factory” initiative, which aims to produce large-scale AI. Along with building large-scale AI infrastructure in South Korea, the company is also preparing new hubs in the Asia-Pacific (APAC) region, where its existing business operations have not yet reached.

Kim Ji-hoon, Director of Naver Cloud, said during a keynote speech at “AI Cloud Big Tech 2026,” held at the L Tower in Yangjae, Seoul, on the 1st, “We are preparing key AI Factory hubs in the Asia-Pacific region, where Naver Cloud’s current coverage has not yet reached,” adding, “Depending on where the AI Factory is established, it will determine the AI hub where technology and corporate investment will converge.”
Kim Ji-hoon, Director of Naver Cloud, explains the current status of AI data center (AIDC) construction and power capacity in key regions, including New York, U.S. (Photo: ReporterShin Yeong-bin )

An “AI Factory” refers to a concept that goes a step beyond a data center simply housing graphics processing units (GPUs). It is an infrastructure that combines data centers, power and cooling infrastructure, high-speed networks, the cloud, and AI models and services to continuously produce AI outputs. Director Kim described it as “core infrastructure for producing intelligence on a massive scale.”

He said, “AI factories meet a nation’s entire AI demand and produce the intelligence needed across the economy, industry, and society,” adding, “Whether a country secures AI factories and who holds the lead in this area will determine its competitiveness.”

From Training to Inference… Growing Demand for AI Infrastructure
Naver Cloud forecasts that
demand
for AI
infrastructure
will grow even more rapidly in the future. While GPU demand was previously concentrated on training large-scale AI models, the company assesses that the proportion of inference—used to run actual services—is increasing.

Director Kim predicted, “While training is used by a handful of AI big tech companies, inference will be used by every company worldwide,” adding, “In the future, inference itself will become a distinct domain of the IT environment, much like the internet or smartphones.” He explained that since AI agents call upon models and tools multiple times even while performing a single task, computational demand increases in tandem with growing service usage.

Physical AI—the application of AI to robots and manufacturing sites—was also identified as a new source of demand. In sectors such as manufacturing and defense, where it is difficult to transmit data and systems externally, there is a need not only for massive central data centers but also for AI data centers built close to the actual sites.

To address this, Naver Cloud plans to build closed-loop clouds within customers’ data centers and provide an environment where GPUs can be operated in a manner similar to a central AI factory. In particular, the company anticipates growing demand for on-site AI infrastructure in sectors where security and data sovereignty are critical, such as defense and manufacturing.

NeoCloud–Hyperscaler Combination
For its AI Factory business, the company also presented a strategy to combine the strengths of “NeoCloud”—which specializes in GPUs—with those of existing large cloud providers.

NeoCloud excels at configuring dedicated AI clusters based on large-scale, state-of-the-art GPUs and high-speed networks. Hyperscalers, on the other hand, possess certification, networks, storage, and a diverse ecosystem of cloud services.

Director Kim stated, “Naver Cloud will advance the AI Factory initiative by leveraging both NeoCloud’s expertise and the competitiveness of hyperscalers.” He highlighted the company’s experience in developing its own AI models and directly operating large-scale GPU infrastructure, data centers, and cloud services as key differentiators.

The target customer base for the AI Factory was also specified. The company plans to supply dedicated infrastructure at the cluster or data center level to global AI companies that utilize large-scale AI infrastructure over the long term, and to provide closed, dedicated environments for sectors with high security requirements, such as national defense. The plan also includes providing AI infrastructure to small and medium-sized businesses, including those using Naver’s own services and Smart Store merchants.

“The scale of AI Factory will continue to expand,” said Director Kim. “The locations where AI Factory is established will become AI hubs, and we aim to turn this market into an opportunity for Korea.”

Economy

Corporation

IT·Science

Economy

Micron Hot on the Heels of SK hynix… DRAM Market Share Gap Narrows to 1 Percentage Point

U.S.-based Micron has increased its market share in the global DRAM market, closing in on second-place SK hynix. The gap in market share between the two companies narrowed from 7 percentage points in …
2026-10-02 13:29:48

Corporation

"Starting on the 6th"... Tension Rises in the Retail and Food Industries Ahead of Parliamentary Hearings

Tension is mounting as officials from major companies in the retail and food industries are being summoned as witnesses ahead of next week’s parliamentary audit. Key industry issues—including personal…
2026-10-02 11:06:55

IT·Science

DOUZONE BIZON to Supply ONE AI PE to Seoul Metropolitan Government’s ‘Agent AI’ Project for the Second Consecutive Year

DOUZONE BIZON(012510)announced on the 2nd that it will supply its generative AI platform, “ONE AI Private Edition (PE),” for the “Administrative Support Agent AI Development” project being promoted by…
2026-10-02 11:37:07