Compute layer
Analysis of AI accelerators, GPUs, CPUs, memory bandwidth, data-center demand, edge computing and the silicon architectures behind compute growth.
This article provides an in-depth analysis of how the surge in demand for AI chips has led to HBM capacity crowding out traditional DRAM and NAND Flash, triggering global supply chain tensions, price increases, and structural changes in the industrial chain, while also looking ahead to future trends and investment opportunities.
The soaring demand for HBM memory in AI data centers has squeezed the DRAM capacity for consumer electronics, leading to a sharp price increase for laptops and iPads. The three major memory manufacturers prioritize the high-profit AI market, causing the price of regular DRAM to surge by approximately 90% within eight months. This structural shift has fueled explosive growth in the refurbished electronics market—Back Market data shows that weekly sales of refurbished MacBooks have increased by 43%. This article analyzes the impact of AI memory allocation on consumer electronics prices from the perspective of the semiconductor supply chain and explains how refurbished retailers have emerged as unexpected winners.
Meta announced that its latest self-developed AI chip will be put into production in September 2026. How will this move affect the AI chip market, the advanced process foundry landscape, and the global semiconductor supply chain? This article provides an in-depth analysis from dimensions such as technology roadmap, industry chain collaboration, competitive dynamics, and regional layout.
Based on the latest Forbes analysis, this article explores the industrial impact of the AI industry shifting from a computing power bottleneck to a system integration bottleneck, analyzing the potential impacts on the chip industry chain, the competitive landscape of cloud platforms, and various regions.
In-depth Analysis of the Profound Impact of Jalapeño, the First Self-developed AI Inference Chip Jointly Released by OpenAI and Broadcom, on the Semiconductor Industry Chain, Technology Roadmap, Market Competition Landscape, and Regional Supply Chain.
Based on a Forbes report, an in-depth analysis of how AI data center construction drives the explosion of memory chip demand, breaks historical cycles, and reshapes the industry chain, competitive landscape, and regional layout.
In the first quarter of 2026, data center IT semiconductor and component revenue increased by 116% year-over-year, driven by AI infrastructure expansion and memory price increases. This article provides an in-depth analysis from the perspectives of the industrial chain, technology roadmap, competitive landscape, and regional impact.
Multiple industry alliances, including automotive, retail, and healthcare, have warned the U.S. government that the surging demand for memory chips from AI data centers is causing supply shortages and price spikes, which could severely impact automobile production. This article analyzes the battle between AI chips and automotive chips from the perspective of the semiconductor supply chain, exploring its far-reaching implications for supply chains, technology roadmaps, and the global competitive landscape.
TechCrunch reports that AI computing demand is shifting from the training stage to the inference stage, driving a reshuffling of specialized inference chips, neoclouds, and data center deployment models. This article analyzes, from the perspectives of chip architecture, advanced packaging, supply chain, and the global competitive landscape, what this change means for NVIDIA, SambaNova, Groq, Cerebras, cloud infrastructure, and the semiconductor industry chain.