Chinese semiconductor industry

Status
Not open for further replies.
D

Deleted member 15949

Guest
Oldschool said:
叶甜春发表

我们最大的失误是2000年代电脑、手机这些半导体设备大量进入家庭，中国成为最大的电子市场的时候没有刻意的培养自己的芯片生产，哪怕就是做点声卡、网上这种最低端但是量大的东西都没有，所以到现在落后太多。

Ye Tianchun

Our biggest mistake was that semiconductor devices such as computers and mobile phones entered the home in large numbers in the 2000s. When China became the largest electronic market, it did not deliberately cultivate its own chip production, even if it was the lowest-end but large-volume things like sound cards and low end Internet devices. Right now, far behind now
Click to expand...
Industry hack asks for money. What else is new?
 
D

Deleted member 15949

Guest
Oldschool said:
Chinese article explains why China early start at lithography fallen behind during opening up in the 90s

Internationalization and globalization are the key culprits. Western firms swarmed into China market. And China ditched their tech independence and focused on labor intensive goods to serve western markets

Chinese people begun to worship western goods and technology and look down upon Indegenious tech.

Foundation laid by early pioneers got wasted.

This is totally different picture than folks spreading falsehood here that China was so backward then got nothing and had to get rich first before moving up.

Had they not totally abandon this and kind keep it on the side.

From 2001 to 2010, China can use that foundation to quickly catch up western tech . 90nm in 2003 would be reachable.

Instead SMEE 90nm came out in 2016, more than 10 years too late.

起了个大早的国产光刻机研发，是如何与行业脱节的？

三次全国大规模集成电路会议，引发国产光刻机成果大爆发

光刻机是大系统、高精尖技术与工程极限高度融合的结晶，被誉为集成电路产业链“皇冠上的明珠”。20世纪50年代，美国研制出接触式光刻机，70年代，美国Perkin-Elmer公司研制成功1:1投影式光刻机，随后美国GCA公司推出“分步重复精缩投影光刻机”，将投影光刻的比例发展到1:5或1:10。分步式光刻机的概念提出之后，光学光刻机的技术路线就基本上稳定下来了，后续的光刻机基本上都属于这种类型。差异只在于光源的变化。

日本的尼康和佳能于20世纪60年代末开始进入光刻机领域。中国利用光刻技术制造集成电路，大致也开始于同一时期。

1965年，我国第一块集成电路在北京、石家庄和上海等地相继问世。上文中提到，1974年9月，第一次全国大规模集成电路工业会议召开，国家计委在北京召开《全国大规模集成电路及基础材料攻关大会战会议》，拟定的目标是【1974～1976年期间，突破大规模集成电路的工艺、装备、基础材料等方面关键技术】四机部组织京沪电子工业会战，进行大规模集成电路及材料、装备研发，突破超微粒干板、光刻胶、超纯净试剂、高纯度气体，磁场偏转电子束镀膜机等材料、装备。



1975年12月，第二次全国大规模集成电路会议在上海召开；1977年1月，第三次全国大规模集成电路会议在贵州召开。这三次会议，可以说直接导致了上世纪80年代前后，中科院系统、电子部系统、地方各研发单位光刻机成果的第一次大爆发。　

根据现存最早的国产光刻机的纪录，是1445所在1974年开始研制，到1977年研制成功的GK-3型半自动光刻机（吴先升.φ75毫米圆片半自动光刻机[J].半导体设备,1979(04):24-28.），这是一台接触式光刻机。1978年，1445所在GK-3的基础上开发了GK-4，把加工圆片直径从50毫米提高到75毫米，自动化程度有所提高，但仍然是接触式光刻机。



GK-3光刻机（刘仲华.GK—3型半自动光刻机工作原理及性能分析[J].半导体设备,1978(03)

1980年左右，中科院半导体所开始研制JK-1型半自动接近式光刻机，于1981年研制成功两台样机，完成第二阶段工艺试验，并进行模拟4K和16K动态随机贮器器件的工艺考核试验。同年，上海光学机械厂的研制的JKG—3型光刻机通过鉴定与设计定型，该机型是我国第一代半自动接近式光刻机。



图：JK-1型光刻机整机照片

由于美国在50年代就已经拥有了接触式光刻机，相比之下中国落后了近二十年。同时国外从1978年开始转向分步重复投影光刻，此时中国科学界也已认识到，分步投影光刻技术的优越性，但限于国内工艺基础差，难以实现。

但是根据八五、九五期间我国微电子技术发展的要求，迫切需要相当数量的分步光刻机，而当时国际上一台i线分步光刻机的售价是160万美元，一台准分子激光DSW光刻机的售价是210万美元，一套g线DSW光刻机也要120万美元，如果全部采用进口设备，当时的国家财力也难以支持。

在此背景下，1978年世界上第一台DSW光刻机问世不久，机电部第45所就开始跟踪研究分步式光刻机，对标美国的4800DSW。1985年，研制出了BG-101分步光刻机样机并通过电子部技术鉴定，认为达到4800DSW的水平。如果资料没有错误，这应当是中国第一台分步投影式光刻机，采用的是436纳米G线光源。

同样在1985年，中国科学院上海光学精密机械研究所研制的"扫描式投影光刻机"通过鉴定，为我国大规模集成电路专用设备填补了一项空白。按照这个时间节点算，中国在分步光刻机上与国外的差距拉近到7年左右（美国是1978年）。

可以看出，国产光刻机研发在上世纪70年代后期起步，直到80年代后期，技术一直在推进，并且取得了一定的代表性成果。

"贸工技"风潮盛行，国产光刻机研发开始脱节？

20世纪50至80年代初期，中国半导体产业蓬勃发展，几乎与世界同时起步，这也是国产光刻机研发的关键产业背景。但是进入80年代中期后，中国却“掉队”了。

此时的日本半导体产业已经发展成为令美国也忌惮的霸主，到了1984年，在光刻机领域尼康和GCA平起平坐，各享三成市占率，Ultratech占约一成，Eaton、P&E、佳能、日立等每家都不到5%。同一年，ASML诞生。

在整个行业积极探索新技术的时候，由于价格高昂和“巴统限制”，“巴统”不批准向我国出口先进设备，国外工艺线已用0.5μm的设备时，却只对我国出口1.5μm的设备，整整差了三代。此外，在80年代，“巴统”规定对我国出口的DSW光刻机，镜头NA必须小于0.17，即只能有2μm以上的分辨率。

也是在这个时期，国内“造不如买”的思想开始盛行，"贸工技"风潮一时盛行，在集成电路等产业也渐渐与国外脱节。产业抛弃了独立自主，自力更生的指导方针，盲目对外开放，中国独立的科研和产业体系被摧毁，研发方面是单打独斗，科研成果转化成商业化产品的微乎其微。

国产光刻机研发的脚步大大减缓。到1990年3月，中科院光电所研制的IOE1010G直接分步重复投影光刻机样机通过评议，工作分辨率1.25微米，主要技术指标接受美国GCA8000型的水平，仅相当于国外80年代中期水平。

这个时期国外光刻光源被卡在193纳米无法更进一步已经长达20年，科学家和产业界一直在探讨超越193纳米的方案，台积电在2002年提出了浸入式193nm技术方案成功解决了这一难题，使得光刻机技术进入到新的阶段。而这时国家科技部才组织实施“十五”863计划“100纳米分辨率193纳米ArF准分子激光器步进扫描投影光刻机”重大项目的研制与攻关，计划在2005年完成试生产样机，2007年小批量生产。

2002年国家在上海组建上海微电子装备有限公司(SMEE)承担“十五”光刻机攻关项目时，中电科45所将从事分步投影光刻机研发任务的团队整体迁至上海参与其中。2008年国家又启动了“02”科技重大专项予以衔接持续攻关。至2016年，上海微电子已经量产90纳米、110纳米和280纳米三种光刻机。

但是，走在世界前沿的ASML已经开始EUV光刻机研发，并于2010年研发出第一台EUV原型机，由于成本过于高昂由三星、台积电、英特尔等公司共同入股推动研发，若干年后，ASML成为7纳米以下制程光刻机的唯一供应商。国产光刻机至此落后ASML超过20年。
Click to expand...
I can't tell if you are trolling that *import substitution* is your magical One Weird Trick to keep China on the IC frontier when even Japanese companies like Nikon and Canon even fell behind the technological frontier in the early 2000s. Sabotaging upstream industries with shittier components is bad actually
 
D

Deleted member 15949

Guest
Oldschool said:
@voyager1
What's your take on this
Mainland media nailed TSMC vile scheme of expanding 28nm just to sabotage SMIC.



突发，台积电露出真面目！花187亿在南京扩产28nm芯片！目的只有一个:打压中芯国际！中国院士当场拆穿台积电谎言

Click to expand...
It's a bad take. You should learn some economics and take a physics class before posting on here to stop your idiocy
 
A

ansy1968

Brigadier
Registered Member
AndrewS said:
Please, Log in or Register to view URLs content!

Please, Log in or Register to view URLs content!
Click to expand...
@AndrewS Bro the Koreans is sounding the alarm together with Mr Morris Chang, we may see a shift and with a US approval stamp on it, the American had gotten what they wanted both the TSMC and Samsung tech. I think all this brouhaha about decoupling is for the US to gain back control. It will be like the western arsenal during WW2 providing Chips for the Western world especially for the 6 eyes. Now let see what Europe will do, will they be a 3rd bloc or join the US, interesting times ahead indeed.
 
horse

horse

Colonel
Registered Member
tsmc-top-customers-2019-2021-jpg.jpg
 
horse

horse

Colonel
Registered Member
ansy1968 said:
@AndrewS Bro the Koreans is sounding the alarm together with Mr Morris Chang, we may see a shift and with a US approval stamp on it, the American had gotten what they wanted both the TSMC and Samsung tech. I think all this brouhaha about decoupling is for the US to gain back control. It will be like the western arsenal during WW2 providing Chips for the Western world especially for the 6 eyes. Now let see what Europe will do, will they be a 3rd bloc or join the US, interesting times ahead indeed.
Click to expand...
Hey brother, check out this old article from 2019.

Please, Log in or Register to view URLs content!


It says the usual about how trade wars will hurt everybody, and there are a couple of charts that worth looking at, and since this guys seems to be an insider to the industry, he makes it a point that a decoupling will damage this worldwide innovation ecosystem.

That last part, I never thought of, but that should be correct.

To simplify things, TSMC makes the chip for Qualcomm who then sells it to a Chinese company like Xiaomi.

But ... what ... if ...

SMIC already at 7nm .... and there is a fully domestic 28nm Chinese fab coming online ... that would mean Chinese would have the know how to make probably 95% of all chips that are needed, once they build capacity in more fabs inside China.

What does that mean for this worldwide innovative ecosystem for semiconductors?

Well, apparently, there will be two such ecosystems, one in China, one outside of China. Both might survive, or maybe just one will survive.

Both of these IC ecosystems will have China as the biggest market by a wide margin, more than double, for the end product, the chip.

----------- ---------- ----------


Now we know what that Huawei executive was talking about in his remarks a couple of weeks ago, that the tech war destroyed worldwide trust.

In other words, if we do not trust you, we will not use your stuff in the long run, and stick to our own tech ecosystem.

This tech war, this round in semiconductors, may be further along in their proceedings than we thought.

:oops:

I mean ... shit, once they start pouring the concrete for a new fab in China, that is the point of no return.

:oops: :D o_O
 
Last edited:
Status
Not open for further replies.
Top