rickawsb

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

Above the semiconductor industry chain, before digital twins: Analysis of the invisible champions improving yield
If we view semiconductor manufacturing as a system, we will find a long-overlooked position: above the industry chain, before digital twins are truly implemented, there exists a layer of cross-enterprise, full-process “cognitive layer” that has not been fully defined. The value of PDF Solutions comes from here.
It deals not with point data, but with the causal chain spanning design, process, equipment, and testing: a certain design structure, forming specific defects at a particula

Recently, the AI chip darling that submitted an IPO—Cerebras—has become a sensation in Silicon Valley.
Its chips, in small model scenarios, can achieve inference speeds up to 20 times that of H100; for ultra-large models (such as 400B parameters), the response speed of the Cerebras CS-3 system for a single user is about 2.4 times that of B200.
So how exactly does Cerebras do it? Will it become a NVIDIA killer?
We need to start from the essence of computing power evolution.
The evolution of AI computing power is shifting from “raw compute” to “communication and system architecture.” On

Weekend Deep Dive: Analyzing the Position, Boundaries, and Endgame of Independent Laser Players from the Trends in CPO + ELS Light Sources
AI computing power bottlenecks are shifting from computation to bandwidth. As GPU scale expands, inter-node communication approaches an N² growth pattern, with electrical interconnects reaching power and distance limits, turning optical interconnects from an "option" into a "necessity."
In this process, CPO (Co-Packaged Optics) and ELS (External Laser Source) are beginning to reshape the industry chain: lasers are being separated from module internals a