Etivus Expands into Physical AI and Robotics

Etivus Group Chairman Jeong Myeong-cheol expects revenue of 2.5 trillion won this year and said the company plans to expand from AI infrastructure into physical AI and the use of robots in workplaces. Etivus is combining systems integration capabilities with partnerships including Dell Technologies, HPE, Oracle, Cisco, VMware, Red Hat and Amazon Web Services to design hybrid AI infrastructure suited to each company’s security and work environment. Jeong said most major corporate and public-sector GPU infrastructure projects in the second half of this year were planned around liquid cooling. He also said adoption of NVL72-based systems, which pack 72 GPUs into a rack, is rapidly increasing among large domestic IT service companies and major business groups.

Market research forecasts South Korea’s GPU server market will grow 7–8% annually to about 7 trillion won by 2029, while Jeong said growth feels above 10% annually in the field. Budgets for individual projects have grown from hundreds of billions to trillions of won, and he expects large investments to continue for at least the next 3 years. He also forecast annual growth of more than 10% for AI data centers, citing power density per rack, cooling standardization and policies to disperse data centers from the Seoul metropolitan area to regional locations as key factors.

Jeong said companies differ in security requirements, investment capacity and existing systems, so not all need to build their own GPU data centers. Etivus proposes on-premises systems (computing equipment kept inside an organization), cloud services or hybrid environments according to customer needs. Its services include GPU server, storage and network design; links to existing business systems; cloud migration; and post-deployment operations and technical support. As Boston Dynamics’ domestic strategic partner, Etivus said it plans to support more than robot hardware supply, working on IT infrastructure, ultra-low-latency communications, real-time data integration and control, and integration with production systems so robots can operate in manufacturing, logistics and hazardous workplaces. It plans an industry-specific robot transition ecosystem spanning pre-deployment consulting, dedicated system construction, operations and maintenance. Hyundai Motor Group’s plan to put Atlas robots into production processes was described as becoming visible in 2028.

Jeong forecast that government and private-sector investment in AX infrastructure began this year and would spread across industries in 2027. He said many companies’ AI infrastructure projects remain at the one-off proof-of-concept or commercialization-preparation stage in actual workplaces, and that success depends on changes to work processes and measurable improvements in productivity, cost and quality. He said that requires coordinated work on internal data preparation, computing and cloud resources, security, and links to existing systems. Etivus aims to use its internally built and validated business generative AI experience and agentic AI technology to provide systems that continue improving after deployment. Jeong said the company aims to become a global, comprehensive AI solutions provider in AI infrastructure, physical AI and AI services in 10 years, and a trusted company sought out first by customers and vendors in 50 years.

Etivus began more than 30 years ago, when Yeongwoo Digital was founded in a roughly one-pyeong office in Yongsan. The company went through the 1997 IMF financial crisis, the venture bubble collapse in the early 2000s, the 2008 global financial crisis and the COVID-19 pandemic in 2020. Jeong named keeping promises to global vendors and domestic customers through changing market conditions as the company’s core value of trust. He graduated from Dongguk University, worked at an insurance company and then founded a business; he also ran Ebraintech. The group now has more than 10 affiliates. Jeong said he chose the name Etivus by combining “Eternal Universe,” which has the same meaning as Yeongwoo.

Etivus Group Chairman Jeong Myeong-cheol expects revenue of 2.5 trillion won this year and said the company plans to expand from AI infrastructure into physical AI and the use of robots in workplaces. Etivus is combining systems integration capabilities with partnerships including Dell Technologies, HPE, Oracle, Cisco, VMware, Red Hat and Amazon Web Services to design hybrid AI infrastructure suited to each company’s security and work environment. Jeong said most major corporate and public-sector GPU infrastructure projects in the second half of this year were planned around liquid cooling. He also said adoption of NVL72-based systems, which pack 72 GPUs into a rack, is rapidly increasing among large domestic IT service companies and major business groups.

Market research forecasts South Korea’s GPU server market will grow 7–8% annually to about 7 trillion won by 2029, while Jeong said growth feels above 10% annually in the field. Budgets for individual projects have grown from hundreds of billions to trillions of won, and he expects large investments to continue for at least the next 3 years. He also forecast annual growth of more than 10% for AI data centers, citing power density per rack, cooling standardization and policies to disperse data centers from the Seoul metropolitan area to regional locations as key factors.

Jeong said companies differ in security requirements, investment capacity and existing systems, so not all need to build their own GPU data centers. Etivus proposes on-premises systems (computing equipment kept inside an organization), cloud services or hybrid environments according to customer needs. Its services include GPU server, storage and network design; links to existing business systems; cloud migration; and post-deployment operations and technical support. As Boston Dynamics’ domestic strategic partner, Etivus said it plans to support more than robot hardware supply, working on IT infrastructure, ultra-low-latency communications, real-time data integration and control, and integration with production systems so robots can operate in manufacturing, logistics and hazardous workplaces. It plans an industry-specific robot transition ecosystem spanning pre-deployment consulting, dedicated system construction, operations and maintenance. Hyundai Motor Group’s plan to put Atlas robots into production processes was described as becoming visible in 2028.

Jeong forecast that government and private-sector investment in AX infrastructure began this year and would spread across industries in 2027. He said many companies’ AI infrastructure projects remain at the one-off proof-of-concept or commercialization-preparation stage in actual workplaces, and that success depends on changes to work processes and measurable improvements in productivity, cost and quality. He said that requires coordinated work on internal data preparation, computing and cloud resources, security, and links to existing systems. Etivus aims to use its internally built and validated business generative AI experience and agentic AI technology to provide systems that continue improving after deployment. Jeong said the company aims to become a global, comprehensive AI solutions provider in AI infrastructure, physical AI and AI services in 10 years, and a trusted company sought out first by customers and vendors in 50 years.

Etivus began more than 30 years ago, when Yeongwoo Digital was founded in a roughly one-pyeong office in Yongsan. The company went through the 1997 IMF financial crisis, the venture bubble collapse in the early 2000s, the 2008 global financial crisis and the COVID-19 pandemic in 2020. Jeong named keeping promises to global vendors and domestic customers through changing market conditions as the company’s core value of trust. He graduated from Dongguk University, worked at an insurance company and then founded a business; he also ran Ebraintech. The group now has more than 10 affiliates. Jeong said he chose the name Etivus by combining “Eternal Universe,” which has the same meaning as Yeongwoo.