Home > Opinion > Columns

Working hours are holding back Korea's chip industry

Published: 04 Mar. 2025, 00:02

Kim Yong-seok

The author is a distinguished professor at Gachon University’s College of Semiconductor.

The amendment to the so-called K-Chips Act, which strengthens tax incentives for semiconductor investments, passed the National Assembly on Feb. 27. However, a provision that would exempt semiconductor research and development (R&D) personnel from the 52-hour workweek regulation was excluded from the final bill.

Korea implemented the 52-hour workweek in July 2018 to address its notoriously long working hours, which ranked second only to Mexico among OECD countries. The intention — to enhance efficiency, reduce excessive labor intensity and improve quality of life — is commendable. However, the law’s rigid, one-size-fits-all application across all industries remains a fundamental flaw.

Industries vary widely, from high-tech enterprises to services and manufacturing, and even within companies, job roles differ — ranging from management and production to research and development. A friend who is an artist once told me, “I worked until 3 a.m. because I was inspired.” Whether in art, technology or product development, innovation thrives on unpredictability. The process of creating something new inherently involves trial, error and persistence.

The development of semiconductor chips follows a two-stage process: prototyping and commercialization (mass production). When news reports hail a “world-first” chip, they often refer only to a prototype. While a prototype may function successfully just once, commercialization is a different challenge — it requires extensive testing, troubleshooting and optimization.

Typically, the process involves design validation, mass production validation and shipment validation. In corporate R&D, both prototyping and commercialization are integral, yet the 52-hour workweek disrupts these workflows. By nature, R&D involves cycles of unpredictable problem-solving. The government and legislators must recognize this reality.

Rep. Kweon Seong-dong, center, floor leader of the conservative People Power Party, speaks at a meeting with government officials to discuss a special law that will exempt semiconductor workers from the country's 52-hour workweek system at the National Assembly in Seoul on Feb. 4. [YONHAP]

High bandwidth memory (HBM) provides a compelling example. SK hynix spent over four years developing HBM1, launching it in December 2013. The HBM design team was so overwhelmed that the division became known as “the wilderness.” When Samsung began mass-producing HBM2 in late 2015, SK hynix faced immense pressure. The market seemed uncertain, but SK hynix persisted, and when OpenAI’s ChatGPT emerged in November 2022, it gained a competitive edge in the HBM market.

By the time SK hynix started supplying HBM3 to Nvidia in June 2022, the company had been working on the technology for over a decade, having initiated the project in 2009. Before the 52-hour workweek took effect in 2017, engineers worked around the clock, dedicating themselves fully to the project — ultimately leading to commercial success.

R&D is inherently unpredictable. Samsung, having underestimated HBM’s potential, is now struggling to catch up. While Samsung’s challenges are unrelated to the 52-hour workweek, SK hynix’s success does not justify rigid time restrictions either.

The most concerning issue is the shift in mindset among researchers. The 52-hour rule has transformed the workplace from goal-oriented to time-oriented. Since its implementation, time has become the target — workers stop as soon as they reach the limit, fostering unproductive habits and inefficiencies.

Of course, researchers should not be forced to work longer hours. The argument is not for compulsory overtime but for flexibility — allowing researchers to work intensively when necessary and rest when needed. The rigid work-hour cap must be lifted. Additional hours should be incentivized with substantial compensation, enabling researchers to experience both fulfillment and tangible rewards for their efforts.

Top researchers do not adhere to fixed schedules. They work when inspiration strikes and rest when needed. Policymakers must understand this. As Donald Trump’s second administration prioritizes America-first policies and Chinese firms like DeepSeek gain momentum, Korea must empower its semiconductor researchers to work without arbitrary constraints. Their competitors in the United States, China and Taiwan face no such limitations. Korea must not be left behind.

Translated using generative AI and edited by Korea JoongAng Daily staff.