TSMC once attempted to apply the principle of shift work to its research and development sector in an effort to close technology gaps with US semiconductor manufacturers. During an event hosted by the United Daily News, Chief Operating Officer Chyn Yung-Pei disclosed details about this initiative, which was internally codenamed Project Nightjar. Although Chyn did not specify the exact timeframe of the program, he confirmed that its objective was to accelerate process development by having internal teams operate for 24 hours a day, compared to the eight-hour shifts typical of US competitors.
The strategy proved unsuccessful, and TSMC declared the project failed after only six months. Chyn summarized the outcome by noting that the efficiency of intellectual work in R&D cannot be increased simply by extending working hours, unlike physical labor. Instead, the company found greater success through a model emphasizing closer cooperation between different departments.
This successful alternative is known as the One Team program. Under this approach, research and development teams collaborate with fab personnel not only during the development of new processes but also for a period following their introduction. Previously, manufacturing processes were developed in laboratories before being handed over to the fabrication plants. Chyn noted that this traditional handover method led to significant problems during the 40-nm process node, prompting the search for new methods.
With the One Team strategy, fab teams no longer have to solve manufacturing issues in isolation; they can now combine their practical experience with the knowledge of development teams. Chyn highlighted that a key challenge is applying this concept globally, as fabs in the US and Japan, along with future facilities in Germany, now operate under the One Team framework. He considers this collaborative model to be the most important pillar for TSMC's manufacturing efficiency and its ability to maintain technological leadership.
Sources: Golem.de.