NHN Cloud sets sights on GPU cloud service with Nvidia’s H100

NHN Cloud Corp., a cloud computing unit of South Korea’s internet company NHN Corp., is setting its sights on a graphics processing unit (GPU)-based cloud service armed with Nvidia Corp.’s H100 chips.

Kim Dong-hoon, chief executive of NHN Cloud, said on Tuesday the company aims to transform into a GPU-based cloud service firm from a central processing unit (CPU)-based company.

“Demand for GPU infrastructure from artificial intelligence companies is rapidly growing. But there is still no absolute leader in the GPU cloud market,” he said.

NHN Cloud recently equipped its data centers in Korea with 1,000 Nvidia H100 AI GPU chips — the largest among Korea’s cloud service providers (CSPs).

NHN Cloud is also building its own AI ecosystem with the launch in 2022 of AI Easy Maker, a platform that allows clients to easily utilize high-performance computing (HPC) cloud servers when developing AI services.

With its cloud services, NHN Cloud plans to soon enter the European and Southeast Asian markets, the CEO said.

KOREA’S BIG 3 CSPs

NHN Cloud’s two other crosstown rivals — Naver Cloud Corp. and KT Cloud Corp. — are also launching AI cloud services.

Last October, Naver Cloud, an affiliate of Korea’s top internet firm Naver Corp., launched Neurocloud for HyperCLOVA X, an AI cloud service optimized for its generative AI model HyperCLOVA X.

KT Cloud, a unit of Korea’s telecom giant KT Corp., recently unveiled AI Train, an AI model learning support service, and plans to launch AIOps, a machine learning analytics tool in the third quarter.

According to market research firm Fortune Business Insights, the global GPU cloud market is forecast to grow from an estimated $4.3 billion this year to $49.8 billion by 2032.

A cloud service is any service available to users on demand via the internet from a cloud computing provider’s server, as opposed to the company’s own on-premises server.

Cloud services are designed to provide easy, scalable access to applications, resources and services, and are fully managed by cloud service providers. Cloud computing is generally used to describe internet data centers or IDCs, a core part of cloud infrastructure.

Write to Dong-jin Hwang at radhwang@hankyung.com
In-Soo Nam edited this article.