'Sharing chipmaking facilities helps cut costs,' Singapore semiconductor expert

Gan, executive director of the Institute of Microelectronics (IME) at Singapore's Agency for Science, Technology and Research (A*STAR), discussed the country's semiconductor development and efforts to build a skilled workforce in a recent interview with VnExpress.

Terence Gan, executive director of the Institute of Microelectronics (IME) at Singapore's Agency for Science, Technology and Research (A*STAR). Photo courtesy of A*STAR

- What role does A*STAR IME play in connecting semiconductor research with manufacturing, and what are its main areas of focus?

- A*STAR IME bridges early research and manufacturing. We take promising solutions that were demonstrated at coupon level and implement them at wafer level using industry-grade tools.

We develop solutions with the following in mind: ease of manufacturing, manufacturing cost, repeatability and reliability. Because of this, our solutions are more ready for commercialization, and hence adoption by the industry is higher.

Our eight R&D focus areas are advanced packaging, silicon photonics, power electronics, RF GaN, piezoMEMS, flat optics, IC design for extreme performance and quantum technologies. These capabilities are used primarily to support Singapore’s semiconductor ecosystem, but we also have customers and partners from many other countries.

We prioritize technologies where the Singapore industry’s needs intersect with Singapore’s strengths and global growth trends. Singapore commits about 1% of GDP to research, innovation and enterprise through its broader RIE plans. Within this commitment, S$800 million (US$625 million) supports the national RIE Flagship in Semiconductors from 2026 to 2030.

We allocate resources to research in our chosen focus areas, upgrading and expanding shared R&D infrastructure, and developing talent. In the short term, we assess how much better our technologies are compared to the current industry benchmarks. Over the long term, we measure how the companies that worked with us translate the technologies into products and services that are delivered from Singapore and contribute to GDP growth.

- How does Singapore connect universities, research institutes and semiconductor companies to turn research into commercial technologies? What has made the biggest difference?

- We increase the rate of successful translation by bringing public researchers from universities and research institutes into national teams. Most research is done as a team effort, but if a member of the team does their research independently, the research is aligned with the rest of the team because the team is guided by a common R&D roadmap. It is also important that the roadmap and research are based on real industry needs.

Our biggest challenge is passing on the knowledge and practical experience that the global industry has accumulated over nearly 80 years, so that young engineers or researchers can contribute effectively. Semiconductor technology moves quickly: between entering university and graduating with a bachelor’s degree, a student could see technology advance by two generations.

Strong scientific and engineering foundations are therefore essential, together with mentorship from research institutes and companies through internships and supervised projects. At A*STAR IME, students work alongside experienced researchers and engineers on real technical challenges. This helps them apply their knowledge, develop practical judgment and keep learning as technology evolves.

- What skills do students need to participate in semiconductor research programs, and how can these opportunities benefit students from Vietnam?

- Research attachments at A*STAR give students opportunities to apply their knowledge and skills to real technical challenges. We welcome PhD students from around the world to spend 1-2 years with us via the A*STAR Research Attachment Programme, or ARAP. To make the most of the research attachments, students need a strong foundation in science and/or engineering, curiosity, the capacity to learn, and good communication skills.

Students need not be from electrical or electronics engineering because semiconductor R&D is multidisciplinary. Ensure the students are strong in their fundamentals. Complement this with exposure to the industry via industrial attachments and/or projects with the industry.

The Singapore Institute of Technology’s (SIT) planned extended Integrated Work Study Programme is one example of this approach. It will give selected students almost two years of workplace learning alongside their studies, with A*STAR IME supporting them through mentorship and hands-on research opportunities.

- Semiconductor research requires expensive equipment, cleanrooms and fabrication facilities. How has Singapore made this infrastructure accessible to universities, researchers and companies without duplicating investments?

- The biggest expense is in industry-grade cleanrooms and equipment. Because Singapore is small, our approach is to invest in one set of such facilities and equipment as a national resource for public research. This also ensures that the fab is better utilized than if each university or research institute has its own fab. It is also important to ensure that the fab is well maintained and used properly.

To do this, the fab is run by a dedicated operations team. Requests to process wafers are subject to review, and once accepted, the recipe is loaded into the manufacturing execution system (MES), and the operations team processes the wafers based on the MES system. This gives different research teams access to the fab through a common, controlled workflow.

It also builds a library of stable processes, allowing researchers to focus on experimental steps while using established processes for the rest. This model helps researchers, universities and companies develop and validate technologies without each organization building a complete fab.

For example, NexGen Wafer Systems, a semiconductor equipment startup, worked with ASTAR IME to develop and validate its equipment faster and cheaper by leveraging ASTAR IME’s fab to process wafers in a fuller flow, with equipment and process expertise that it does not have. For a startup, reducing upfront investment in infrastructure frees up resources for product development and business growth.

From Singapore’s experience, research does not happen in isolation. The best researchers are those with industry and manufacturing experience. Singapore began demonstrating our ability to do good R&D after building up decades of experience in manufacturing. Our talent development efforts are successful because of stable industrial policies, sustained investments in education and industry growth. To paraphrase, people must be interested in investing in their own development because they see this as a way to a better future for themselves and their family.

Singapore’s strengths in semiconductors are the result of almost 60 years of industrial development. Today, we are strong in wafer fabrication of specialty ICs and NAND flash, advanced packaging, and final testing. Many of the top semiconductor companies also do R&D in Singapore. In fact, the electronics industry is the largest contributor to business expenditure on R&D and has one of the highest value-added per worker in manufacturing in Singapore.

- After nearly six decades of semiconductor development, what lessons can Singapore share with Vietnam, and where could the two countries complement each other?

- Singapore’s experience shows that building a semiconductor industry takes time, patience and persistence. Stable policies that support business investment give companies confidence to commit to long-term projects. Sustained links between research, manufacturing and talent development also matter.

Companies bring technical challenges; researchers develop and test possible solutions; and universities help prepare people who can carry the work forward.

The semiconductor supply chain already connects capabilities across countries. Rather than assign Singapore and Vietnam fixed roles, I would look for areas where companies and research teams have a shared need and complementary strengths.

That gives any collaboration a clear purpose and value for both ecosystems. To do good translational research, it is important to listen to the needs of the industry in Vietnam and build multi-skilled innovation teams.