China’s startup, DeepSeek, has unveiled a low-cost open-source AI model, igniting optimism in the market about potential reductions in AI costs and advancements in edge computing. The company’s continued open-source initiatives have been met with positive feedback from developers, underscoring growing enthusiasm for accessible AI tools. However, Ubitus’s CEO, Wesley Kuo caution that while DeepSeek’s large model holds promise, it remains better suited for cloud-based applications at this stage. The company still faces significant challenges in the edge AI space, where established models from competitors continue to hold sway.
The demand for automotive lenses appears to be on an upward trajectory, while machine vision and drone applications are simultaneously capturing widespread attention. In this context, manufacturers of automotive lens modules assert that the demand for drone lenses will not only match but may even surpass that of robotics. This market, characterized by small-batch, high-variety production, is particularly conducive to fostering product and technological differentiation.
Benefiting from the booming demand for testing related to artificial intelligence (AI), CCP Contact Probes, a probe development and manufacturing company, has optimized its product structure and is making strides in the semiconductor and electric vehicle (EV) sectors. Recently, positive news has emerged regarding test sockets and probes, which are expected to penetrate the advanced packaging supply chain for AI chips.
US President Donald Trump, now back in the White House, has once again set his sights on TSMC. Previously, his administration championed the semiconductor supply chain crisis as a national security issue, urging the Taiwanese chip giant to establish manufacturing in the US. Now, the stakes have risen—Trump is threatening to freeze CHIPS Act subsidies, impose new semiconductor taxes, and, in an even more audacious move, reportedly demand that TSMC step in to save Intel.
Google has launched a free version of Gemini Code Assist, providing developers with access to AI-powered code completion. The offering includes 180,000 code completions per month, a significant increase compared to the 2,000 completions offered by Google’s previous free tools. This introduction positions Google to compete with existing AI coding platforms, notably Microsoft’s GitHub Copilot.
As global interest in unmanned vehicles expands beyond aerial drones, the development of surface vessels and underwater unmanned vehicles (UUVs) is gaining increased attention. In the past several years, countries like the US and China have invested substantial resources in enhancing fully autonomous and semi-autonomous platforms for both surface and underwater operations. Given Taiwan’s geographic position surrounded by water, industry experts emphasize that developing UUVs is essential, despite development hurdles being significantly greater than those faced in aerial drone systems.
Taiwan’s renewable energy ambitions are facing mounting hurdles, as regulatory and logistical challenges slow the rollout of solar photovoltaic (PV) and offshore wind projects. The Ministry of Economic Affairs (MOEA) has lowered its renewable energy generation target from 20% to 15% by the end of 2025. However, even this reduced goal appears increasingly unattainable, with renewables accounting for just 12% of Taiwan’s total electricity generation as of 2024.
With non-Chinese markets entering their traditional off-season and the Chinese market lacking a nationwide shopping promotion event, overall smartphone shipments by Chinese brands are expected to decrease by 9.7% on quarter in the first quarter of 2025 after enjoying good sequential growth in the fourth quarter of 2024, according to DIGITIMES’ latest report covering China’s smartphone industry.