LG Innotek is accelerating its shift toward AI-driven growth, combining semiconductor substrates for AI infrastructure with sensing and robotics components as it seeks to reduce dependence on its traditional mobile-focused businesses.
LG Innotek is reshaping its business around two technology markets gaining momentum from artificial intelligence: semiconductor substrates for computing infrastructure and physical AI components for autonomous machines.
The South Korean electronics components manufacturer reported first-half revenue of $8.22 billion and operating profit of $401.96 million, according to the company, representing year-over-year increases of 24% and 296%, respectively. The performance is notable because the first half is traditionally an off-peak period for the business.
The results also illustrate the company’s broader portfolio strategy under CEO Moon Hyuksoo, who took the role in 2023. LG Innotek has been reallocating resources toward higher-value businesses designed to provide more stable earnings while building new sources of growth.
One of those areas is semiconductor substrates, which connect and support semiconductor packages in increasingly complex computing systems. The technology has become more strategically important as AI accelerators, servers and memory systems demand greater bandwidth and increasingly sophisticated packaging.
LG Innotek has decades of experience in the substrate business and has used that expertise to expand beyond mobile applications. Its RF-SiP substrates, for example, use copper-post technology and have maintained a leading position among the company’s five largest global customers since 2016, according to LG Innotek.
The company’s semiconductor substrate portfolio is also expanding into AI-related computing. FC-CSP substrates, historically associated with mobile application processors, are gaining applications in memory products as AI infrastructure drives demand.
Meanwhile, LG Innotek has expanded its FC-BGA business into computing. After beginning mass production of FC-BGA products for PC chipsets last year, the company says it began high-volume production for PC CPUs in the third quarter and plans to enter AI accelerator and AI server markets sequentially from next year.
That transition is strategically important because AI computing is increasing demand for advanced packaging across processors, accelerators and memory. Semiconductor packaging has become a critical part of system performance as chip designers push toward higher compute density and bandwidth.
LG Innotek expects its package solution business to become a larger contributor to earnings. The company has set a target of $743 million in operating profit from its package solution business, including semiconductor substrates, by 2031.
The second part of the strategy moves beyond the data center and into physical AI.
LG Innotek is applying its expertise in optical components to autonomous driving and robotics. The company is developing a portfolio covering automotive cameras, LiDAR and radar, giving it multiple sensing technologies for machines that need to perceive and respond to physical environments.
Autonomous driving is currently one of the company’s more advanced physical AI commercialization opportunities. In March, LG Innotek announced strategic partnerships including one with U.S. autonomous-driving software company Applied Intuition to develop autonomous-driving solutions.
Robotics is another emerging target. LG Innotek signed an agreement with Boston Dynamics last year to develop robotic components and more recently announced a strategic partnership with Japan’s TDK covering sensing technologies for physical AI, including vision modules and tactile sensing modules for robotic hands.
The company says its robotics component business generated revenue for the first time this year and that it is already mass-producing camera modules for humanoid robot customers at its Paju facility in South Korea.
The opportunity extends beyond cameras. Humanoid robots require multiple technology layers, including perception, communications, power and semiconductor components. LG Innotek’s existing expertise across optical modules and semiconductor substrates gives it a potential route to supplying several of those components.
That strategy mirrors a broader evolution in AI hardware. Generative AI has concentrated investment around GPUs and data centers, but the next wave of AI applications is increasingly moving into vehicles, robots and other physical systems. Those applications require a combination of sensors, edge computing, communications and specialized components rather than compute alone.
LG Innotek is consequently positioning itself as a supplier across both sides of that AI infrastructure transition. Semiconductor substrates connect the company to the expanding AI computing ecosystem, while cameras, LiDAR, radar and tactile sensing provide exposure to physical AI.
The company says it plans to grow future businesses, including semiconductor substrates and physical AI solutions, to $5.94 billion in revenue by 2030, representing 25% of total revenue.
Whether those targets are achieved will depend on AI server demand, semiconductor packaging cycles, autonomous-driving adoption and the pace at which humanoid robotics moves from early deployments to larger-scale production.
For LG Innotek, however, the portfolio shift is already visible. The company is using its existing manufacturing and component expertise to build a business that reaches from AI servers and accelerators to autonomous vehicles and robots.
That makes its AI strategy less about developing an AI model and more about supplying the physical technology required to make AI systems operate in both digital and real-world environments.
Market Landscape
AI infrastructure is expanding beyond GPUs and data centers into advanced semiconductor packaging, sensors and robotics. As AI accelerators require increasingly complex system architectures, packaging technologies such as FC-BGA and FC-CSP are becoming strategically important.
At the same time, physical AI is creating demand for cameras, LiDAR, radar, tactile sensing and other perception technologies. Companies with established expertise in optical and semiconductor components are seeking opportunities across autonomous vehicles and robotics.
LG Innotek’s strategy combines these trends, using semiconductor substrates to target AI computing while extending optical and sensing capabilities into physical AI.
Top Insights
- LG Innotek is expanding semiconductor substrates from mobile applications toward PC processors, AI accelerators and AI server infrastructure.
- Its physical AI strategy combines cameras, LiDAR, radar and emerging tactile sensing technologies for autonomous vehicles and robotics.
- Partnerships with Applied Intuition, Boston Dynamics and TDK indicate an effort to build an ecosystem around autonomous systems and robotic components.
- Humanoid robotics gives LG Innotek an opportunity to reuse optical-component expertise while expanding into new sensing and connectivity requirements.
- The company expects future businesses, including semiconductor substrates and physical AI solutions, to reach $5.94 billion in revenue by 2030.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
