Chinese semiconductor industry

Status
Not open for further replies.

vincent

Grumpy Old Man
Staff member
Moderator - World Affairs
I think posted the by the infamous Huawei whisperer. Given the lack of future of ARM in China, I see it really as a fools errand for Hisilicon to keep designing smartphone CPUs around ARM. I wish they devote more effort with RISC-V or LoongArch. Since domestic process for advanced CPU won't be ready until probably 2025, they got some time to start trying out RISC type of architecture. Huawei should know better than anyone to not trust in ARM.
My amateur take: Huawei has an architecture license from ARM, that means they design SOCs that interprets and execute ARM instruction sets instead of using cores designed by ARM. How the Huawei SOC executes the instructions is somewhat instruction-independent. It is probably not too difficult for Huawei to switch the core design to RISC-V.
 

hvpc

Junior Member
Registered Member
I believe there are many second hand DUVi in the market.
@PopularScience, in your opinion, what is considered as “many”?
Would the second hand DUVi be excluded from the potential ban ?
If the potential sanction involves DUVi, it’s likely one could still buy the used tool, but the problem would be how to service these refurbished DUVi systems.

the very limited refurbished DUVi transaction in recent years still go through ASML/Nikon for the installation and service. If new sanction limits ASML/NIKON to sell their DUVi, it would limit them to service it as well.
 

PopularScience

Junior Member
Registered Member
@PopularScience, in your opinion, what is considered as “many”?

If the potential sanction involves DUVi, it’s likely one could still buy the used tool, but the problem would be how to service these refurbished DUVi systems.

the very limited refurbished DUVi transaction in recent years still go through ASML/Nikon for the installation and service. If new sanction limits ASML/NIKON to sell their DUVi, it would limit them to service it as well.

enough to fill up 2 to 3 14nm fabs.

Please, Log in or Register to view URLs content!
 

ansy1968

Brigadier
Registered Member
From my mentor @Oldschool, how many university and State scientific institution are involve in China IC development, aside from Peking University, CAS and Tsinghua, we also have Harbin Institute of Technology (HIT) working on Integrated Circuit and Lithography machine light source, who else?

Please, Log in or Register to view URLs content!




Many friends asked how the light source of the extreme ultraviolet EUV lithography machine developed by Harbin Institute of Technology is progressing ?


Oldschool

Junior Member​

Registered Member
15 minutes agoNew
Peking University involved in both DUV and EUV development.

3装备先进一体化光刻协同设计研制技术北京理工大学

2002年12月至今,在国际自然基金委、“十五”期间科技部重大项目、“十一五”——“十三五”国家重大专项资助下,带领团队率先利用光刻仿真软件PROPLITH,研究分析光刻机性能参数、掩模结构及参数、工艺参数对光刻成像性能的影响,为光刻机研制企业提供研究分析数据。率先建立了NA1.35浸没光刻机全光路、非傍轴近似、非远心、含3D偏振像差的严格矢量成像理论和仿真软件,实现更高精度的正向光刻仿真分析;同时,率先建立的严格矢量计算光刻、全曝光视场-低误差敏感度-多目标计算光刻、压缩感知快速计算光刻、高保真-贝叶斯压缩感知、全芯片-快速计算光刻技术等多种先进、快速、高保真、自适应计算光刻等分辨率增增强技术。面向未来9-1nm节点IC制造EUV光刻机发展的趋势和需求,建立了EUV光刻物镜的分组设计与优化等方法,设计了缩小倍率4、5和8,及数值孔径NA0.25、0.3、0.33、0.4、0.5、0.55、0.6和0.7的四反、六反、八反、十反一系列结构的非球面和自由曲面、共轴或离轴的EUVL物镜及其相匹配的照明系统,并对多层膜与物镜进行协同光学设计研究,获得高分辨、反射率均匀的EUV光刻成像设计结果;同时也对部分物镜系统进行了热变形分析与像差补偿设计、可制造性约束及其性能分析,设计了多种结构的一系列物镜及与之匹配的照明系统。其中NA0.33 EUV曝光系统,能够满足22~13nm光刻技术节点设计性能的需求。同时也较早地开展了NA0.33 EUV 计算光刻研究。2016年春季率先设计并发表NA0.5-0.6变倍率EUV物镜,2017年,首次发表了NA0.5-0.6变倍率物镜需要的EUV光刻机照明系统设计,可在掩模面实现95%~98%的照明均匀性;同时对高NA大口径曝光系统需要的高性能多层反射膜进行了创新型设计,并申请了多项专利保护。

4470f592c78ff17c72daa1b1a753164e.png@wm_2,t_55m+5a625Y+3L+WMl+S6rOWmh+iBlA==,fc_ffffff,ff_U2ltSGVp,sz_27,x_17,y_17


近期,在高NA0.55 EUV物镜和照明系统设计以及计算光刻方面,取得了新的进展。在2019年,申请人及团队提出了一种用于EUV的基于梯度的计算光刻分辨率增强方法,研究了厚掩模低偏振像敏感度的计算光刻。研究经历和成果表明,她在DUV和EUV光刻机的照明系统-成像系统设计、加工、检测及其与掩模和工艺之间的协同设计优化方面,独树一帜。

为了验证光刻极协同设计研制技术,2008年开始,团队自主设计并研制了NA0.75/小视场DUV研究型光刻机、NA0.5反射式-类EUV光刻机,并在此研究型光刻系统集成了多种像质检测传感器(图1-2,图3a),实测2个研究型小视场光刻成像性能与协同设计研制预期相符,证明了光刻设备与掩模/工艺/检测的一体化协同设计优化和研制技术的有效性。创新性建立了多种先进计算光刻方法(图3b)。引领和支撑企业和研究院所DUV、EUV光刻机研制。在光刻机偏振像差检测技术基础上,研制了新型-多功能偏振成像/光谱/OCT技术与系统(图4-9)。相关技术也将有利地支撑光刻机偏振器件和偏振像差高精度检测,并可拓展到半导体材料与器件、超颖表面特性表征、以及航天/遥感/化工/医药、医学诊疗等领域。

96a82e6f47e5abad9e15e694133cdf66.png@wm_2,t_55m+5a625Y+3L+WMl+S6rOWmh+iBlA==,fc_ffffff,ff_U2ltSGVp,sz_26,x_17,y_17


c1942bb1695d3b765fc5de4976c5c6f5.png@wm_2,t_55m+5a625Y+3L+WMl+S6rOWmh+iBlA==,fc_ffffff,ff_U2ltSGVp,sz_16,x_10,y_10



1a3237dbff3e1fa48eae5c6c08e82484.png@wm_2,t_55m+5a625Y+3L+WMl+S6rOWmh+iBlA==,fc_ffffff,ff_U2ltSGVp,sz_24,x_15,y_15


Like Quote Reply
Report
 

european_guy

Junior Member
Registered Member
@tonyget, do you think the iLine price tag is too high for a SSB800?

SSB800 at no less than 50M RMB seems quite expensive for a front end iLine system. But it’s not a surprise the system price is so high considering what UPrecision is reportedly charging for the DWS stage.

With the current exchange rate, you could get a Canon FPA5550 iZ2 for <25M RMB that’s at least 50% more productive. Even ASML iLine could be had at ~35M RMB.

What is panning out is an interesting anomaly.

Usually a newcomer starts with the entry level and gradually goes up to advanced level machines.

But because of

1. Price/productivity of SMEE is not competitive at entry level i-line machines

2. If there will be a ban, it will be for immersion litho, not for mature nodes

3. ICRD is testing the Arf dry version for a reason. it is more critical than the i-line or the Krf

By 2024 we could have just a small bunch of SMEE i-line machines in production (or at least assumed as such), that mainly are used to debug the software and the machine's general functionality and reliability...and this requires a lot of time to polish.

But the real numbers of SMEE machines will be in the dry Arf and later in the immersion Arf, where they will grow faster...inheriting all the software improvements that in the meantime are debugged out on the i-line machines.

In 2/3 years time we could have in the market many more SMEE Arf and Arfi machines than the lower level Krf and i-line. That would be quite odd but is a real possibility.
 

tonyget

Senior Member
Registered Member
A clearer answer from havok. There is no 65nm machine,only 28nm and 90nm

havok: 上微没有65nm光刻机,至少现在没有。只有28和90
havok: SMEE have no 65nm lithography machine, at least not now. only 28nm and 90nm machine

It's good to know that transition from 28nm to 14nm will be faster(than transition from 28nm to 90nm)

华粉:有没有可能我国从28nm到14nm的时间反而快点?
havok:嗯
huafen:Is it possible that China's transition from 28nm to 14nm will be faster?
havok:Yes
 

tphuang

Lieutenant General
Staff member
Super Moderator
VIP Professional
Registered Member
This week's bid results.
Please, Log in or Register to view URLs content!
25 won by mainland
6 by Taiwan
40 by Germany
4 by US
3 by Netherlands
4 by Japan
12 by korea

25 out of 84 items for domestic firms.
so this is an odd week where most bids were own by international bidders, especially by a couple of German vendors.
 

tonyget

Senior Member
Registered Member
This week's bid results.
Please, Log in or Register to view URLs content!
25 won by mainland
6 by Taiwan
40 by Germany
4 by US
3 by Netherlands
4 by Japan
12 by korea

25 out of 84 items for domestic firms.
so this is an odd week where most bids were own by international bidders, especially by a couple of German vendors.

Where is 25 come from?So far I only count 18 won by mainland firms

810710763337.6919.jpg
 
Status
Not open for further replies.
Top