rickawsb

The Performance Competition of MLCC in the AI Data Center Era: Why Are Murata and Taiyo Yuden Significantly Ahead?
AI data centers are driving MLCC into a new cycle of technological upgrades.
In the past, servers mainly used 12V power supplies; now, they are evolving toward 48V rack power, and in the future, even into the 800V HVDC high-voltage direct current era. Meanwhile, AI platforms like NVIDIA GB200 and GB300 are continuously increasing in power consumption, with GPU core voltages dropping to 0.6V-0.8V, yet single GPUs are drawing over 1000A of current.
For MLCC, the main challenges come

Over the past six months, a significant change in the industry has been that advanced packaging has shifted from supporting roles to core functions for the first time.
HBM, CoWoS, ABF substrates, high-speed interconnects, power delivery, and advanced packaging capabilities are increasingly becoming bottlenecks in the supply chain.
Because AI chips are changing rapidly. Die sizes are getting larger, more HBM stacks, more chiplets, higher power consumption, and increased thermal density.
As a result, the complexity of each chip's packaging is beginning to rise non-linearly.
Advanced pack

Understanding the Next Big Military Race in Power Semiconductors Under the AI Data Center Cycle, Beyond GPUs, It's Power
AI data centers are getting larger and larger, with energy consumption of a single data center often comparable to a medium-sized city.
In the past, data centers used 10-20kW per rack, now it's 80kW, 120kW, or even 600kW per rack. Power consumption of large AI clusters has entered the GW level.
The bottlenecks are no longer just GPUs, CPUs, and storage, but also current, heat, power distribution, copper losses, power conversion efficiency, grid access, and HVDC.
AI d

The booming demand in the AI industry chain has driven storage to explode over the past six months, pushing low-end phones out of the market.
In the next six months, it will drive the explosion of CPUs, pushing low-end computers out of the market.
Currently, it is driving the explosion of batteries and power semiconductors, and in another six months, it will also push low-end new energy vehicles out of the market.

The end of AI is the lithography machine, and the end of lithography machines is the lens
--- Why are EUV lithography lenses so difficult?
EUV and high-end DUV optics are the ultimate convergence of the ultra-precision industrial system. They depend on materials, coatings, metrology, assembly and adjustment, thermal control, vibration control, algorithms, error modeling, and long-term experience accumulation. The real bottleneck to expanding production is often not a single component, but the entire “precision closed loop.”
The core of this loop is: you cannot manufacture something more precis

Weekend Industry Research — Analyzing the Market Share of SiC, GaN, and Silicon MOSFETs in AI-Driven Power Electronics Infrastructure
The rapid construction of AI data centers is driving a major upgrade of the power grid, bringing another long-underestimated field back into the spotlight: power semiconductors.
The core of the power system is efficiently controlling current. The most essential device for current control is the MOSFET (Metal-Oxide-Semiconductor Field-Effect Transistor).
Over the past few decades, almost all global power devices have been built on silicon MOSFETs. Silicon i