Micron Technology is turning the AI memory boom into a new Wall Street story: not just record DRAM, NAND and HBM demand, but stronger free cash flow, long-term customer commitments and a clearer path to shareholder returns.
Micron Technology expects the global memory market to remain supply-constrained beyond 2027, driven by surging artificial intelligence demand and structural limits on new manufacturing capacity, while unveiling a new long-term supply model backed by strategic customer agreements.
The global semiconductor packaging and testing race is heating up as TSMC recently signed a 10-year agreement with Amkor to expand advanced packaging collaboration in Arizona, drawing close market attention to how the rivalry with ASE and Amkor will reshape market share. ASE Chief Operating Officer Tien Wu said he is "optimistic," stating that strong demand from US customers and the need to diversify supply-chain risk make deeper global investment inevitable.
Co-packaged optics (CPO) is rapidly emerging as a foundational architecture for next-generation AI infrastructure. With Nvidia's latest Scale-Up CPO switch roadmap now taking shape, bandwidth per AI rack is set to increase from approximately 130 TB/s in the Blackwell generation to more than 1 PB/s in the Feynman era, spanning Blackwell, Rubin, Rubin Ultra, and Feynman platforms. As a result, the global AI race is rapidly expanding beyond GPU compute, with optical interconnects, silicon photonics, and CPO emerging as the next major battleground.
JCET Group is accelerating its AI chip packaging expansion with a CNY7.8 billion (US$1.1 billion) investment in a new advanced packaging and testing plant in Shanghai Lingang, strengthening capacity for AI computing, high-performance chips, high-density memory and automotive electronics.
Global helium supply is under renewed strain. Nippon Sanso, Japan's largest industrial gas supplier, announced it will raise prices across its helium product line by an average of more than 30% starting July 2026, citing persistent tightness in global supply driven in part by rising geopolitical risks in the Middle East.
As demand for AI chips surges, the battle for foundry orders is heating up. South Korean industry sources say the next 2-3 years will be a critical period for Samsung Electronics' foundry business, with its 2028 turnaround target hinging on stable operations at the Taylor plant and landing major orders.
Rising demand for AI server builds is tightening the global supply of high-end MLCCs, driving steep price increases in China's electronics distribution market and exposing how quickly infrastructure demand can affect component costs worldwide. The rally underscores a structural mismatch that could keep pressure on buyers across technology supply chains for years, according to industry sources.
ByteDance's reported plan to purchase at least 50,000 AI inference chips from Shanghai-based GPU developer Iluvatar CoreX could become one of the most significant commercial wins yet for China's domestic AI chip industry.
IC design firms are scrambling to expand their partner networks as tight packaging and testing capacity remain the most pressing bottleneck in the supply chain. Companies in Taiwan, Europe, and the US say the shortage is worse than expected, pushing packaging, testing, and related materials into broader supply constraints.
At Nvidia's June 25 annual shareholder meeting, CEO Jensen Huang declared that national security "comes first" wherever it conflicts with commercial opportunity, pledging full compliance with US export controls while casting the chipmaker as a core pillar of America's AI and semiconductor industrial base.
More coverage
