Chinese semiconductor industry


krautmeister

Junior Member
Registered Member
From that tidbit Chinese news, we only know cetc finished design the 35nm duv. And if it's dual stage then it leapfrog smee and it can go for 14nm at least

@WTAN, @foofy to keep an eye on this cetc duv.

Also euv and duv are parallel. Euv can come out in 2025. Who knows?
Why would CETC dual stage lithograph indicate they are ahead of SMEE in 14nm? This would only indicate their lithograph was more productive, but not necessarily ahead in process.

If SMEE alignment accuracy is indeed not on par with CETC, that is a different story. From what I know, SMEE immersion optics are stuck at NA 0.75 but there are rumors of new NA 1.35 optics in the works. News concerning patterning techniques like SADP are all rumors and a black box to the public at this point.
 

Oldschool

Junior Member
Registered Member
Why would CETC dual stage lithograph indicate they are ahead of SMEE in 14nm? This would only indicate their lithograph was more productive, but not necessarily ahead in process.

If SMEE alignment accuracy is indeed not on par with CETC, that is a different story. From what I know, SMEE immersion optics are stuck at NA 0.75 but there are rumors of new NA 1.35 optics in the works. News concerning patterning techniques like SADP are all rumors and a black box to the public at this point.
Dual stage is must have for finfet. Non finfet can rely on single stage.
All the optics are from guo hwang. From Chinese forum, there seems confirmation NA 1.35 accomplished
Smee may opt for easier overall design at 28nm.
 
Last edited:

Oldschool

Junior Member
Registered Member
首台浸润系统样机很早就交付了。

整个项目不止28nm节点,还有55-40nm节点,如果投产的话这些成熟节点比28的快,上微干式双台光刻机进度也快于浸没式。


看来之前交付3台浸润系统应该是进行量产验证了 2021

干式机双工件台在北京的产线施工已有部分产品从华卓亦庄园区发货

SSA800 还没有验收

燕东微电子 90nm

这个2022年完工的全国产验证线是28nm节点,不是14nm。14nm是研发中心三期的项目。
还有28nm光刻机的镜头已经好了但还没发运,出厂仪式在节后举行。

这条线建设地点在嘉定工业区,占地13万平米,是月投片2万片的12英寸集成电路芯片生产线,28nm节点,属于国家级项目。采用国产光刻机(包括800系列的多种型号),已准备开建了。

NA1.35的镜头已经完工了.本来昨天就要发运的

机应该交给icrd,后面还可能交付给2022年完工的嘉定工业区的全国产验证线。另外北京也会建一条大的全国产验证线,也在排队等着样机。smic想买,估计需要等完全量产才行。而ssa800所用到的华卓精科的双工台,也至少要等到2023年了,因为该双工台的量产是华卓精科亦庄二期来做,而该二期产线计划明年上半年才开工


北方华创的铜互联pvd肯定到不了14nm……这一块是amat垄断的,已经发展了至少两代工艺

涂胶显影设备现目前只有芯源微在搞,现目前只能满足晶圆前道28nm工艺节点及以上工艺制程,满足14nm的还在研制

南大光电现在开发的是28nm节点的光刻胶,14nm/7nm工艺的要另外开发而且难度更大
 

SleepyStudent

Junior Member
Registered Member
From that tidbit Chinese news, we only know cetc finished design the 35nm duv. And if it's dual stage then it leapfrog smee and it can go for 14nm at least

@WTAN, @foofy to keep an eye on this cetc duv.

Also euv and duv are parallel. Euv can come out in 2025. Who knows?
So if I'm reading this correctly, the estimates for the 3 projects in parallel are
ArF Immersion (1.35NA), Single Stage: 2022/2023
ArF Immersion (1.35NA), Dual Stage: 2024
EUV (0.33NA): ~2025
 

gelgoog

Captain
Registered Member
That is taking way too long. Some of it was to be expected since you can't rush these things too much. But the market window to manufacture DUV might close if EUV improves. China needs to have a parallel development program which not only is working on ArF and first generation EUV but also later generation EUV. I also think that China needs to come up with something to break the market. One thing the Chinese government could do would be to fund an initiative to manufacture with 450mm wafers ahead of everyone else. China has enough volume they could drive this change and it would enable China to have massively more production capacity than anyone else.
 

krautmeister

Junior Member
Registered Member
Dual stage is must have for finfet. Non finfet can rely on single stage.
All the optics are from guo hwang. From Chinese forum, there seems confirmation NA 1.35 accomplished
Smee may opt for easier overall design at 28nm.
Given NA 1.35 optics coupled with the 28nm SMEE ArF 193nm DUV immersion lithograph, this would still not result in 14nm capability without the requisite overlay precision. Dual wafer stages is primarily to increase throughput by splitting the wafer alignment and exposure. Various patterning schemes all rely in 1 way or another on overlay precision. If overlay precision can be improved, then even 7nm is doable with SMEE 28nm lithograph with NA 1.35.

From what I know, the lack of high NA lens (>1.0) plus relatively poor overlay precision are the 2 remaining bottlenecks preventing <=14nm capability. I'm talking only about the lithograph, not other things like lack of full flow EDA, sub-28nm photoresists, sub-28nm ion implanters, etc.
 

SleepyStudent

Junior Member
Registered Member
Given NA 1.35 optics coupled with the 28nm SMEE ArF 193nm DUV immersion lithograph, this would still not result in 14nm capability without the requisite overlay precision. Dual wafer stages is primarily to increase throughput by splitting the wafer alignment and exposure. Various patterning schemes all rely in 1 way or another on overlay precision. If overlay precision can be improved, then even 7nm is doable with SMEE 28nm lithograph with NA 1.35.

From what I know, the lack of high NA lens (>1.0) plus relatively poor overlay precision are the 2 remaining bottlenecks preventing <=14nm capability. I'm talking only about the lithograph, not other things like lack of full flow EDA, sub-28nm photoresists, sub-28nm ion implanters, etc.
Yeah, the question is what is the NA and overlay precision of the SSA800. I thought it was 1.3NA with dual overlay but apparently @Oldschool claims it isnt
 

krautmeister

Junior Member
Registered Member
That is taking way too long. Some of it was to be expected since you can't rush these things too much. But the market window to manufacture DUV might close if EUV improves. China needs to have a parallel development program which not only is working on ArF and first generation EUV but also later generation EUV. I also think that China needs to come up with something to break the market. One thing the Chinese government could do would be to fund an initiative to manufacture with 450mm wafers ahead of everyone else. China has enough volume they could drive this change and it would enable China to have massively more production capacity than anyone else.
The CAS recently confirmed next gen SSMB EUV light source was adequate for industrialization. SSMB EUV light source is a higher power source of EUV than LPP, think many hundreds of watts. Problem is, it's only now being industrialized so it can still be 5+ years away.
 

Top