Samsung Electro-Mechanics and LG Innotek, both South Korean electronics component makers, are rapidly expanding the technologies they have accumulated in mobile and IT components into AI servers, autonomous driving, and humanoid robots. The two companies are significantly increasing R&D investment while simultaneously applying AI to development and production sites to boost competitiveness across their businesses.

According to South Korea’s Financial Supervisory Service Electronic Disclosure System on the 20th, Samsung Electro-Mechanics and LG Innotek’s combined R&D expenses for the first half of this year totaled 881.5 billion won (approximately $631.9 million), up 18.2% from 746 billion won in the same period last year. Simply annualizing the current trend would put the full-year figure above 1.76 trillion won (approximately $1.3 billion).

Samsung Electro-Mechanics’ first-half R&D expenses came to 474.1 billion won (approximately $339.8 million), up 26.0% from 376.2 billion won a year earlier. Its R&D-to-revenue ratio also rose 0.3 percentage points from 6.8% to 7.1%, against revenue of 6.6663 trillion won. The company is focusing on developing high-value products including ultra-high-capacity multilayer ceramic capacitors (MLCCs) for AI servers and data centers, flip-chip ball grid array (FC-BGA) substrates for AI accelerators, and camera modules for autonomous driving. It is also developing ultra-high-voltage MLCCs for electric vehicle powertrains in parallel.

LG Innotek invested 407.3 billion won (approximately $292.0 million) in R&D during the first half, up 10.2% from 369.8 billion won a year earlier. While its R&D-to-revenue ratio declined from 4.1% to 3.7% as revenue (11.0621 trillion won) grew at a faster pace, the investment amount itself posted double-digit growth. The company is cultivating next-generation substrates and advanced packaging for AI servers and high-performance computing (HPC), composite sensing for autonomous driving, and sensing and control solutions for humanoid robots as new growth engines.

The two companies’ first-half R&D investment status is as follows:

CategoryR&D Expenses (₩100M)YoY Growth RateR&D-to-Revenue RatioSamsung Electro-Mechanics4,74126.0%7.1%LG Innotek4,07310.2%3.7%Total8,81518.2%-

Note: Based on South Korea’s Financial Supervisory Service Electronic Disclosure System, first-half 2026 results

Beyond supplying AI components, both companies are also applying AI to their R&D and production processes. Samsung Electro-Mechanics operates an “AI Solution Team” and an “Agentic AI Development Task Force” dedicated to AI-based development and manufacturing efficiency. LG Innotek uses AI to predict process defects and derive optimal production recipes, and has applied AI vision to raw material incoming inspections, reducing defect cause analysis time by 90% compared to the previous approach.

The R&D expansion is seen as an extension of strategies to diversify business structures that had been concentrated on smartphone components such as camera modules. Samsung Electro-Mechanics is broadening the applications of its MLCCs and package substrates into AI servers, data centers, and automotive electronics, while LG Innotek is extending technologies secured in camera modules and substrates into AI, semiconductors, autonomous driving, and robotics. The plan is to simultaneously expand both the application areas of existing core technologies and the customer base.

The cash reserves held by both companies also provide a financial foundation for sustaining mid-to-long-term R&D. As of end-June this year, Samsung Electro-Mechanics and LG Innotek held cash and cash equivalents of 3.3144 trillion won (approximately $2.4 billion) and 1.2404 trillion won (approximately $889.2 million), respectively, totaling more than 4.5 trillion won (approximately $3.2 billion).

The expanded investment is also expected to translate into improved earnings. Samsung Electro-Mechanics anticipates record third-quarter results, driven by demand for AI data center and autonomous driving components. KB Securities noted that LG Innotek’s AI semiconductor substrate utilization rate is approaching 100% and projected that second-half operating profit will reach its highest level since the second half of 2021.

Components supplied to AI servers, data centers, automobiles, and robots require high performance and reliability. Since considerable time is needed from component development through customer evaluation to mass production, completing technology validation before the market expands in earnest is critical to gaining an edge in subsequent order competition. An industry insider said, “The AI and physical AI markets demand high levels of performance and reliability from customers, and product evolution is rapid. How quickly a company completes technology validation through advanced R&D will determine its future competitiveness in winning orders.”