Chinese semiconductor thread II

tokenanalyst

Brigadier
Registered Member

Annual output of 100,000 tons! New developments in Shaanxi high-purity industrial silicon project​


It is reported that the project is invested by Shangnan Zhongjian Industrial Co., Ltd. with nearly 1 billion yuan. It plans to build 8 25.5MVA industrial silicon electric furnace production lines, which mainly include main production facilities, auxiliary production facilities, public works, living welfare facilities, etc. After completion, it will produce 100,000 tons of high-purity industrial silicon annually, and achieve an annual output value of 2 billion yuan. The construction period of the 100,000 tons of high-purity industrial silicon project is from 2023 to 2025, and it will be implemented in two phases. It is expected that the 8 25.5MW high-purity silicon electric furnace production lines will be completed and put into production by the end of 2025.​
1736440404911.png

Please, Log in or Register to view URLs content!
 

vincent

Grumpy Old Man
Staff member
Moderator - World Affairs
Yes, this is an entirely ridiculous policy.

View attachment 142961

The DC blobs have gone off on the deep end and fully decided that AI chips are like military weapons.

so this probably restricts the legal sales of Nvidia GPUs (even the low end ones) to China to a certain fixed amount. I presume Huawei is thanking them for this since it has plenty of Ascend NPUs to sell.

But other than that, I'd assume the smuggling business through Japan, Korea, Singapore & Taiwan will continue to bring in plenty of Nvidia GPUs.

It does open up Ascend sales to non-US allied countries.
Higher Resolution:
1736441605659.png
 

tokenanalyst

Brigadier
Registered Member

Study on Tin Contamination Deposition Mechanism on Collector Mirror Surfaces in LPP-EUV Light Sources Based on Hydrodynamics and Particle Simulation Theory.​

Abstract​

The surface of the laser-produced plasma EUV light source collector mirror can accumulate tin contamination during operation. This contamination reduces reflectivity and disrupts the normal functioning of lithography machines. However, the working environment of the EUV light source is extremely complex, making it difficult to clarify the mechanisms of tin contamination deposition. To address this challenge, this study employs fluid dynamics and collisional radiative theory to investigate the generation mechanisms of high-energy tin plasma and to clarify the spatial distribution characteristics of tin plasma in different charge states. A model of the tin plasma expansion process reaching the collector surface was developed using the particle-in-cell model and Monte Carlo methods, examining the tin contamination deposition patterns under various conditions. The results show that tin deposition on the mirror collector is uneven, with significant spatial variation, and an average deposition rate of 2.20×10⁻⁵ nm/pulse. This study is significant for understanding the deposition mechanism of tin contamination on collector mirror and exploring effective methods to mitigate tin contamination.​

Highlights​


  • This study uses fluid dynamics and collisional radiative theory to examine the generation of tin plasma and its spatial distribution across different charge states.

  • A PIC-MCC model was developed to simulate the expansion process of tin plasma reaching the collector surface.

  • The deposition patterns of tin contamination on the collector surface under different conditions were investigated.

  • Tin deposition on the collector mirror is uneven, with significant spatial variation, and an average deposition rate of 2.20×10⁻⁵ nm/pulse.

    Please, Log in or Register to view URLs content!

 

Eventine

Junior Member
Registered Member
This really isn't surprising when you consider how "all-in" the US is on AI, and how deeply the establishment - however tech. ignorant they might be - have been convinced by "AI influencers" like Sam Altman, Eric Schmidt, and Peter Thiel of the impending AI-driven 4th Industrial Revolution. The transformative potential is being hyped to the moon, for example just today an article was released on CNN about how CEOs are predicting 40% of the global work force will be replaced by AI by 2030.

US elites see dominance of the 4th Industrial Revolution as critical to the project for the New American Century. It's really their last chance at retaining hegemony over the world system. They will do anything to win and China is their biggest rival in AI - the only other country that even poses a threat.

Chips (& chip making equipment), data centers (& data), and software AI platforms (& models) are perceived as the three pillars of AI power. It is not an accident that the US is moving to embargo China on all three fronts.
 

tokenanalyst

Brigadier
Registered Member
This really isn't surprising when you consider how "all-in" the US is on AI, and how deeply the establishment - however tech. ignorant they might be - have been convinced by "AI influencers" like Sam Altman, Eric Schmidt, and Peter Thiel of the impending AI-driven 4th Industrial Revolution. The transformative potential is being hyped to the moon, for example just today an article was released on CNN about how CEOs are predicting 40% of the global work force will be replaced by AI by 2030.

US elites see dominance of the 4th Industrial Revolution as critical to the project for the New American Century. It's really their last chance at retaining hegemony over the world system. They will do anything to win and China is their biggest rival in AI - the only other country that even poses a threat.

Chips (& chip making equipment), data centers (& data), and software AI platforms (& models) are perceived as the three pillars of AI power. It is not an accident that the US is moving to embargo China on all three fronts.
Is DJI all over again. US companies with a head start in the technology become popular, the stooges in DOD saw military potential in the technology without exactly defining that potential, DC stooges put exports controls allowing sales to a few selected countries, a small Shenzhen startup started to fill that gap and at the end become the dominant player.
 

tokenanalyst

Brigadier
Registered Member

Sub-nano misalignment sensing for lithography with Talbot self-interference​

Abstract​

In lithography, the sensing of subnanoscale misalignment is deemed impossible for existing alignment systems due to signal loss caused by the optical diffraction limit. To solve this fundamental problem, this paper proposes a misalignment sensing strategy based on Talbot self-interference. This strategy improves measurement sensitivity by using the interference of diffracted optical field of moiré fringes. Experimental results show that the proposed strategy can achieve real time-lapse misalignment sensing with an accuracy of 0.25 nm. We hope that this strategy can provide an effective solution to subnanoscale misalignment sensing, as well as inform other fine-scale measurement approaches.​

Please, Log in or Register to view URLs content!
 
Top