CONNECT WITH US
Jul 23, 09:52
In-depth: Alibaba builds three-layer AI empire across chips, memory and models
Competition in China's AI industry is moving beyond large language models into chips, memory and computing infrastructure. CXMT's planned IPO on Shanghai's STAR Market has put renewed attention on Alibaba, the Chinese DRAM maker's largest industry investor.
STMicroelectronics (ST) has raised its revenue ambition for AI data centers to above US$1 billion in 2026 and well above US$2 billion in 2027, staking a larger claim on the AI infrastructure buildout as the European chipmaker swung back to profit in the second quarter of 2026.

Samsung Electronics is preparing a die-to-wafer hybrid bonding production line with about 50 bonders at its Pyeongtaek campus for next-generation high-bandwidth memory and logic chips, according to The Elec.

The AI build-out has found its next pinch point. After spreading from Nvidia's accelerators to memory and networking gear, the scramble for capacity has now reached the once-plentiful server processor, and Chinese cloud and internet firms are being pushed into the kind of long-term supply commitments once reserved for the tightest parts of the market.

Texas Instruments' second quarter of 2026 results landed Wednesday with two signals worth parsing beyond the headline numbers: an automotive demand acceleration that TI's own management didn't see coming three months ago, and the first analog price increases the company has executed in years. Together, they suggest the current upcycle is broadening in ways the supply chain has not fully absorbed.

Participation in the Global Pavilion at the upcoming SEMICON Taiwan 2026 has reached a new high, with the number of participating countries climbing to 18 and the booth count surging from 62 in 2023 to 220 in 2026, an increase of more than 250%. Both metrics set new records, reflecting that amid rising artificial intelligence (AI) investment and the regionalization of supply chains, countries are moving faster to deepen cooperation with Taiwan's semiconductor industry.

Alphabet's Q2 earnings call reveals a company doubling down on compute capacity, custom silicon, and frontier model development — while investors probe the limits of the strategy

China's research and academic community appears to be drifting away from international collaboration, even as the UK, Germany, and the US have steadily expanded cross-border partnerships in recent years. The contrast has drawn fresh attention after Moonshot AI's flagship 2026 model Kimi K3 and other Chinese technology products, including Unitree Robotics's H1 robot, prompted questions over how much of China's advanced innovation is truly homegrown.

Apple plans to introduce a new device leasing program, marking a significant departure from how it has previously sold its products. The new payments scheme may help convince customers to keep investing in hardware at a time when rising component costs are forcing smartphone companies to raise their prices.

Lightelligence is positioning optical interconnects, optical switching, and optical computing as the next foundation of AI infrastructure, arguing that the industry's biggest challenge has shifted from faster chips to moving data efficiently across increasingly larger GPU clusters.

Samsung is in talks to invest about EUR1 billion (approx. US$1.14 billion) in French AI startup Mistral as part of a wider fundraising round, the Financial Times reported, citing people familiar with the matter.

As the global AI race intensifies, Germany's Merck is extending its reach into semiconductor materials research with an "AI for AI" strategy. Using its in-house Elvis AI platform, BaBE Bayesian optimization tool, digital twin, and enterprise knowledge platform Melinda, the company is building a closed-loop innovation model in which AI drives materials development and new materials, in turn, support next-generation AI chips.