HBM– tag –
-
Semiconductor Back-End
Micron Invests $9.3 Billion in Hiroshima for HBM Production Launching 2028, Expands Packaging Capacity in Singapore
TechsCurrent Japan Overview Micron Technology has broken ground on a $9.3 billion expansion of its Hiroshima, Japan factory to bolster High-Bandwidth Memory (HBM) production, with shipments expected around summer 2028, supported by Japan... -
Semiconductor Back-End
AI Memory Shortage Sparks Advanced Packaging Crisis: Ultrafast Lasers and Glass Substrates Emerge as New Bottlenecks
Chanelink China Overview A 'capacity crisis' in advanced packaging processes, triggered by the AI memory shortage, is causing ultrafast laser suppliers to become a new bottleneck in the supply chain. AI accelerators demand highly advance... -
Semiconductor Back-End
HBM and ABF Shortages Intensify AI Chip Supply Bottlenecks, Ajinomoto Forecasts Over 20% ABF Supply Gap by 2027
Semiconductor Engineering USA Overview The primary bottlenecks for AI chip production are the critical shortage of High-Bandwidth Memory (HBM) and the complex packaging capacity required for GPUs/XPUs by Nvidia, AMD, and Google, coupled ... -
Semiconductor Back-End
TSMC Significantly Expands CoWoS Advanced Packaging Capacity, to Begin Production of 14-Reticle-Size Packages by 2028 for Surging AI Demand
Biya Global Limited Taiwan Overview TSMC is substantially expanding its CoWoS (Chip-on-Wafer-on-Substrate) advanced packaging capacity to meet the surging demand for AI chips. The company aims to nearly quadruple CoWoS capacity to 130,00... -
Semiconductor Back-End
South Korean Conglomerates Samsung and SK Group Invest $252.5 Billion in Chungcheong Province, Bolstering HBM Packaging and AI Semiconductor Supply Chain
The Korea Herald South Korea Overview South Korean giants including Samsung Electronics, SK hynix, and Celltrion announced a combined 392 trillion won (approximately $252.5 billion) investment in Chungcheong Province across semiconductor... -
Semiconductor Back-End
Nvidia Kicks Off Volume Production of Blackwell AI Accelerators with TSMC’s CoWoS-L, Supply to Remain Tight Through Q4
MetalSemi Asia Taiwan Overview Nvidia has commenced volume production of its next-generation Blackwell AI accelerators at TSMC, leveraging the 4NP process and advanced CoWoS-L packaging. Cloud service provider demand is projected to outs... -
Semiconductor Back-End
Lam Research Drives AI Performance with Advanced Packaging, Tackling Bottlenecks and Boosting Power Efficiency through Chiplets and Panel-Based Manufacturing
Lam Research USA Overview Lam Research highlights advanced packaging as the critical innovation frontier for AI, addressing the limitations of traditional semiconductor scaling. By enabling chiplets to function as a single system, this a... -
New Technology
TrendForce Reports Samsung Files New HBM Dummy Die Patent to Enhance 16-Layer HBM5 Stack Reliability
TrendForce Taiwan Overview TrendForce reports that Samsung has filed a new HBM dummy die patent aimed at enhancing 16-layer HBM5 stack reliability. This patent seeks to address warpage and heat dissipation issues in high-stack configurat... -
New Technology
SK Hynix’s Advanced MR-MUF Packaging Process Reduces HBM4 Thermal Resistance by 17%, Contributing to AI Accelerator Chip Stabilization
Kynix China Overview SK Hynix's Advanced MR-MUF (Mass Reflow Molded Underfill) packaging process has been shown to reduce HBM4 thermal resistance by 17% compared to conventional HBM4. This innovative technology is crucial for maintaining... -
New Technology
Kynix Explains HBM4 Hybrid Copper Bonding Achieves Over 20% Thermal Performance Improvement, Enabling Stable Clock Speeds Under AI Loads
Kynix China Overview Kynix explains that Hybrid Copper Bonding (HCB) technology for HBM4 improves thermal performance by over 20%. This method eliminates solder bumps and underfill, directly connecting copper layers to significantly redu...