CONNECT WITH US
Sign out

Taiwan approves funding for local IC design houses to advance AI and chip innovation

, Taipei
0

Credit: DIGITIMES

As Taiwan seeks to strengthen its position in the global semiconductor race, the government is shifting its focus toward homegrown innovation. Taiwan's Ministry of Economic Affairs (MOEA) has approved NT$3 billion (US$97.69 million) in funding for five semiconductor research and development projects involving six domestic companies, with the government providing NT$840 million in subsidies. The initiative, targeting an estimated industrial output value of NT$98.1 billion, represents a strategic pivot toward supporting local IC design houses rather than foreign chipmakers, particularly in cutting-edge areas like artificial intelligence and high-performance computing.

The approved projects include major players such as Phison Electronics Corp., Syntronix, Powerchip Semiconductor Manufacturing Corporation (PSMC), Himax Technologies, Rehear Audiology, and AuthenX. These companies plan to develop advanced semiconductor technologies with an emphasis on artificial intelligence (AI), high-performance computing, and next-generation communications infrastructure.

Taiwan's shift to homegrown chip innovation

This funding effort is part of the MOEA's "IC design topping subsidy program," which aims to prioritize local firms in semiconductor innovation over foreign-dominated beneficiaries. Previously, firms like Nvidia and Micron, both foreign companies, had been the major recipients of government subsidies. To address this imbalance, the program focuses on nurturing Taiwan's homegrown IC design capabilities, particularly in high-value fields such as AI and compute-intensive applications.

In 2024 alone, the ministry approved 11 projects under this program, including significant contributions from MediaTek and Rehear Audiology targeting 5nm process technology. Startups like Neuchips and InPsytech have also made strides with niche products. This reflects Taiwan's commitment to building a diversified semiconductor ecosystem with a balance between established players and emerging innovators.

The MOEA has outlined four priority sectors for 2025 subsidy eligibility: AI, high-performance computing, automotive electronics, and next-generation communications. Submissions are encouraged to incorporate AI-driven hardware-software integration, advancing both industrial and consumer-facing technologies.

AI-driven projects target efficiency and performance

Phison Electronics recently completed a 6nm AI solid-state drive storage controller in 2024 and now seeks to develop its next-generation 6nm AI controller chip. The company aims to create an integrated AI inference microservice platform combining hardware and software for enhanced performance.

Syntronix and PSMC have jointly proposed developing indium gallium zinc oxide (IGZO) analog memory featuring compute-in-memory functions alongside 3D memory stacking. These technologies target the reduction of power consumption, a crucial factor in AI applications where energy efficiency remains a challenge.

Himax Technologies will pursue an ultra-low-power edge AI vision computing platform that integrates image processing chips with infrared sensors. This initiative targets markets such as AI-enabled personal computers, smart wearable devices, security systems, and smart retail environments.

Rehear Audiology plans to enhance its Bluetooth chip technology to create a self-adjusting hearing aid embedded with six AI models capable of automated hearing tests, noise reduction, gain compensation, and real-ear measurements.

AuthenX focuses on silicon photonic chip development that integrates laser sources with advanced packaging technology, aiming for transmission speeds up to 3.2 terabits per second (Tb/s) in line with co-packaged optics (CPO) standards. The goal is to improve interconnect performance for AI graphics processing units (GPUs) and foster integration into the global semiconductor supply chain.

Article translated by Jingyue Hsiao and edited by Jerry Chen