Singapore Launches ‘SG Semiconductor’ as AI Push Moves Chip Race beyond Manufacturing

Singapore has launched a unified national brand for its semiconductor industry, bringing research, manufacturing, infrastructure and talent under a single ‘SG Semiconductor’ banner as the city-state vies for a larger share of the global AI chip boom.

Unveiled on Friday (September 25) by Tan See Leng, minister for trade and industry, the initiative was developed jointly by the Agency for Science, Technology and Research (A*STAR) and the Economic Development Board (EDB). The brand will initially champion Singapore’s public-sector research capabilities across seven foundational areas, including advanced packaging, silicon photonics, power electronics and integrated circuit design, among others.

Although the branding is new, the strategy behind it goes well beyond a routine marketing exercise. Singapore already produces around 10% of the world’s chips and 20% of global semiconductor manufacturing equipment. The sector accounts for nearly 6% of the city-state’s GDP and employs more than 35,000 people. Between 2022 and 2025 alone, the country secured more than S$30 Bn in semiconductor investment, according to EDB data.

Singapore’s immediate focus is to close the critical gap between laboratory research and large-scale manufacturing by localising the development of new chip technologies, proving their commercial viability and seamlessly transitioning them into high-volume production.

This bridge is vital as the generative AI hardware race reshapes semiconductor performance. While leading-edge logic chip manufacturing remains concentrated in Taiwan and South Korea, AI architectures increasingly rely on advanced packaging, chiplets and high-speed optical links to allow processors and memory components to communicate efficiently.

To capitalise on this shift, Singapore is deploying S$800 Mn into a flagship semiconductor R&D programme under its Research, Innovation and Enterprise 2030 framework, focussing explicitly on advanced packaging and photonics. Furthermore, a separate S$500 Mn National Semiconductor Translation and Innovation Centre (NSTIC) is slated to open in 2027, allowing start-ups and SMEs shared access to industry-grade 12-inch tools and cleanroom infrastructure. The overarching goal is to compress the timeline from laboratory to fabrication plant.

A wave of strategic industry partnerships accompanied Friday’s launch to accelerate this commercialisation pipeline. Notably, Applied Materials and A*STAR are entering the fourth phase of a research collaboration dating back to 2011. This five-year expansion will scale up infrastructure, equipment and headcount at their joint laboratory, focussing heavily on hybrid bonding for 3D chiplet integration, high-density interposers and advanced substrates tailored for AI workloads.

The partnership has already yielded more than 10 semiconductor manufacturing tools developed entirely in Singapore. In the past five years, it has connected A*STAR with as many as 50 companies, including nine local SMEs, backed by S$46 Mn in industry funding.

Applied Materials is also scaling its local physical footprint. In June 2026, the US equipment giant opened a $500 Mn (S$600 Mn) campus in Tampines, more than doubling its advanced cleanroom capacity to meet surging global demand for AI chip-production hardware.

Similarly, GlobalFoundries is deepening its silicon photonics partnership with A*STAR, developing technologies on 300 mm wafers for high-performance computing and data communications. The US-based foundry has steadily anchored its presence here, acquiring Singapore-based Advanced Micro Foundry in 2025 to fold an established silicon-photonics business into its local operations.

Silicon photonics has emerged as a critical battleground as AI data centres grapple with the extreme latency, power consumption and thermal thresholds generated by massive data transfers between processors. Leveraging light rather than electrical signals offers a viable route to breaking this bandwidth bottleneck.

Rounding out the alliance, US process-control specialist KLA will kick off a three-year collaboration with A*STAR’s Institute of Microelectronics (IME) to detect packaging defects earlier and boost manufacturing yields.

Meanwhile, assembly and test provider STATS ChipPAC will license IME-developed advanced packaging technology for co-packaged optics, targeting high-volume production for AI and data centre infrastructure.

Together, these projects outline SG Semiconductor’s core thesis. Of course, Singapore is not attempting to clone Taiwan’s dominance in leading-edge logic foundries or South Korea’s monopoly on memory chips. Instead, its strategy targets the exact frontiers where semiconductor performance depends on integration, packaging and specialised manufacturing at scale.

While A*STAR identifies advanced packaging and silicon photonics as Singapore’s primary pillars, the broader semiconductor roadmap extends to power electronics, radio-frequency gallium nitride, flat optics and specialised sensors.

This specialised focus lowers the barrier to entry for agile start-ups and localised suppliers. By offering shared open-access research infrastructure, Singapore mitigates the prohibitive capital expenditure that traditionally prevents a lab prototype from being tested on production-grade machinery.

Ultimately, this strategy’s success hinges on execution. Publicly funded research generates rich intellectual property, but transitioning it to global markets requires massive capital injection, niche tooling and early market alignment. The government’s thesis is to forge those commercial connections much earlier in the cycle.

“SG Semiconductor provides a gateway to Singapore’s ecosystem, enabling companies to develop, scale and commercialise next-generation technologies from here,” said Jermaine Loy, managing director of the EDB, speaking at the launch.

The macroeconomic timing is highly favourable. While the generative AI boom drives computing demand to unprecedented heights, it is simultaneously forcing global chipmakers to look past traditional transistor scaling for efficiency gains. For Singapore, the opportunity is no longer just about manufacturing more silicon; it is about intellectual ownership. The SG Semiconductor label will ultimately be judged by how effectively its homegrown laboratory technologies find their way into the bedrock of the global technology supply chain.