i find a page that cross reference with above
上海微电子、沈阳芯源（芯源微）、盛美、北方华创、中微公司、沈阳拓荆、沈阳中科（中科仪）、成都南科、华海清科、北京中科信、上海凯世通（万业企业）、中科飞测、上海睿励、上海精测（精测电子）、科益虹源、中科晶源。
Chinese semiconductor industry
- Thread starter Hendrik_2000
- Start date
- Status
上海微电子、沈阳芯源（芯源微）、盛美、北方华创、中微公司、沈阳拓荆、沈阳中科（中科仪）、成都南科、华海清科、北京中科信、上海凯世通（万业企业）、中科飞测、上海睿励、上海精测（精测电子）、科益虹源、中科晶源。
If you look at pattern on chip, it's a physical layout of structure of antenna, both transmitter and receiver structure, consists of integrated waveguide, inductor, capacitor. Its RF chip leveraging the cmos process. Its an analog chip based on normally digital chip process.@WTAN @Oldschool looks like a 28nm chips to me. The chip was released by the China Electronics Technology Group Corporation No.38 Institute.
rom CnTechPost
China-made millimeter wave chip sets world record for detection distance
February 19, 2021
A Chinese research institute released a high-performance 77 GHz millimeter wave chip and module on February 17, with a detection distance of 38.5 meters, setting a new global record for the longest detection distance of a millimeter-wave packaged antenna.
The chip was released by the China Electronics Technology Group Corporation No.38 Institute in 2021 International Solid-State Circuits Virtual Conference ( ISSCC), and for the first time, two 3-transmitter, 4-receiver millimeter-wave chips, and 10-way millimeter-wave antennas were integrated in a single package, according to the .
It implements a multi-channel millimeter-wave radar transceiver front-end in a 24 mm × 24 mm space and creatively proposes a dynamically tunable fast broadband chirp signal generation method.
The chip also uses multi-feed antenna technology in the package to substantially increase the effective radiation distance of the package antenna, providing a small size and low-cost solution for proximity smart sensing, the report said.
It uses a CMOS process to integrate three transmit channels, four receive channels, and radar waveform generation in a single chip, which has special advantages in terms of fast broadband radar signal generation and supports multi-chip cascading and building larger radar arrays.
Fan-out-based wafer-level packaging is a mainstream implementation of packaged antennas, and major international companies have developed chip products with integrated packaged antennas based on this technology.
The institute will further optimize the millimeter-wave radar chip and change it according to the expansion of application requirements and technological advances to provide a one-stop solution for specific application scenarios.
ISSCC is considered the "Olympics" in the field of integrated circuits and was founded in 1953 by Bell Labs, which invented the transistor, and other organizations. In its 60-year history, many milestone inventions in the history of integrated circuits have debuted here
Its a front end physical structure chip, it will need baseband chip to process the information.. maybe a CETC dsp chip.
Overall,it's a nice piece to have.
Its for car application.
Its not an uncommon application. Coin size microwave chips are commonly used with conformal arrays, flat panel SATCOMs, wifi routers, phones, etc,.
The chips here below are enormous due to being an S-band application, the antenna needs to be physically half the size of the wavelength. Larger die size also means more power to the chip.
Last edited:
Who's roadmap is that? 150W at IF vs at-source is an order of magnitude difference.
There is another technology similar to DPP, called LDP that is even more compact than DPP. The Japanese company Ushio, that acquired the Dutch-German Xtreme Technologies is now selling a LDP source for EUV mask inspections. Xtreme technologies LDP source was installed in a few ASML EUV machines at customer sites, but they were only able to produce 35W at IF. LDP proved to be very difficult to scale to higher powers, as the plasma heat was compromising the electrodes, whereas with LPP the plasma temperature and intensity did not pose a problem and could be scaled indefinitely by making the CO2 laser more powerful. The CO2 laser used by ASML's EUV systems is more powerful than some military lasers!
According to Insiders, CAS has not yet made a decision on which light source to use, DPP or LPP. Harbin Inst is developing DPP and Shanghai Inst is developing LPP.
There is now a competition going on between DPP and LPP. There is a Beijing and Shanghai race going on. They might end up developing a EUV with 2 types of Light Source. SMEE and CETC might both develop their own EUVLs.
Harbin Institute DPP light source has source power of 150W+. It is still under development and they will certainly try to achieve at least 200W, but there is no news so far about this currently.
Apparently DPP appears to be the favoured option due to cost factors and IP reasons?
If CAS goes ahead to develop a EUVL using the 150W+ DPP Light Source, they will have a product in 2025 which is somewhere in between the NXE3350 (125W) and NXE3400 (200-250W).
This should just be sufficient for some form of mass production.
CAS is still working on the Optics but i think it should be similar in resolution to that of the ASML EUVLs.
It is good news that Huawei has expressed interest in the EUVL that CAS is developing. Even better that CAS has promised to deliver a product soon.
Companies like Huawei and SMIC will be the ones that will help with the Commercialisation of the CAS EUVL.
If they were so shortsighted and had no planning there wouldn’t even be an EUV instrument developed in research.
Hendrik_2000
Lieutenant General
CAS is government agency I am not talking about government agency They do have the foresight and for years exhorted the private companies to be self sufficient but they didn't listen so the EUV stuck as university or research project It does not commercialize because nobody interested in funding it for further development. What they should do is to form semiconductor consortium to fund and provide the seed money and management to take over the EUV project from CAS and developed it into commercial product like they do now but it is late by 15 years
Its from Xtreme.
Actually Only IF power is needed. At source EUV power go through debris migration, collector, and spectral purity filter and finally becomes IF power which is the final useful or clean light.
So, thats why at source power is much bigger than the final clean light, or IF power.
You can compare Huawei and ZTE to gain inside to their mindsets,
Huawei is acting now because it got fully sanctioned and has no choice albeit a bit too late . Now it forming its fabs , getting into semiconductor manufacturing process development and supporting research institutes.
On the other hand, ZTE thinks it home free now after paying US $1billion USD fine. So it continue to do the old style, designing its own chip and fab out to TSMC. I have not heard any news that ZTE is actively doing things that Huawei is currently doing.
chinese industrial players are mostly risk aversion. They try to avoid as much risky projects and investment as possible.
Industrial players participation is critical. ASML wouldn't survive without the continuous support from Intel, Samsung, TSMC. There were doubts about their research on EUV at early periods.
Last edited:
- Status