Chinese semiconductor thread II

tokenanalyst

Brigadier
Registered Member
Interesting patent from AMEC for an integrated etching-deposition tool for making ultra high aspect ratio holes.

Wafer processing method and integrated etch deposition device for wafer processing​

WO2024125303A1​

Abstract​

The present invention discloses a wafer processing method and an integrated etching deposition device for wafer processing. The method comprises: providing a plasma processing apparatus; providing a wafer to be processed and placing said wafer in a plasma processing chamber; starting a plasma radio frequency source and introducing an etching gas to etch an etching area, so as to form a hole or a trench; and performing monitoring and, according to a monitoring condition, performing a change to introduce a first mask precursor to deposit a mask or introduce the etching gas to continue etching until it is detected that the hole or trench formed in the etching area meets a target etching requirement, wherein the first mask precursor contains methylsilane as a substitute. In the present invention, by means of real-time monitoring, the operation mode of wafer processing can be changed to an etching mode or a mask deposition mode, so that in situ etching or in situ mask deposition is performed on a wafer to be processed, and it is unnecessary to remove said wafer from a plasma processing chamber, thereby reducing costs and shortening the production period. The present invention is applicable to the etching of a large-size silicon hole, and is also applicable to the etching of a microstructure of a high or ultra-high aspect ratio.​

1724429025694.png1724429213892.png
1724429248676.png
 

LuzinskiJ

Junior Member
Registered Member
Once all these fab expansions are over SMIC will have similar capacity to TSMC in 7nm and higher. Hua Hong will have similar capacity to UMC in 28nm and higher. Nexchip will have similar capacity to Powerchip.

So basically China will replicate whatever fabrication Taiwan has, except SMIC won't be able to do the EUV fabrication at 5nm and lower done at TSMC, Hua Hong does not have a 14nm FinFET process, Nexchip still needs a 28nm process.
Sanctions on Chinese Semi is like squeezing a balloon…
 

tokenanalyst

Brigadier
Registered Member

The first domestic silicon carbide ingot laser stripping equipment was put into production​

On August 21, news came from Jiangsu General Semiconductor Co., Ltd. that the first domestic 8-inch silicon carbide ingot laser fully automatic stripping equipment independently developed by the company was officially delivered to Guangzhou Nansha Wafer Semiconductor Technology Co., Ltd., a leading company in the field of silicon carbide substrate production, and put into production.



W020240823352743382671.jpg


The equipment can realize fully automatic slicing of 6-inch and 8-inch silicon carbide ingots, including ingot loading, ingot grinding, laser cutting, wafer separation and wafer collection. It fills the market gap in the research and development and manufacturing of domestic silicon carbide ingot laser stripping equipment, breaks through foreign technology blockades, and will greatly enhance the level of independence and industrialization of my country's silicon carbide chip industry.

The equipment can peel off 20,000 silicon carbide substrates per year with a yield of over 95%. Compared with the traditional wire cutting process, it greatly reduces product loss, and the price of the equipment is only 1/3 of similar foreign products.

In recent years, the share of silicon carbide power devices in the high-power semiconductor market has continued to increase, and they have been widely used in a series of fields such as new energy vehicles, urban rail transit, wind power generation, high-speed mobility, and the Internet of Things.

However, due to the high hardness and brittleness of the material, when using traditional cold cutting processes such as mortar wire and diamond wire to cut and peel silicon carbide ingots, there are disadvantages of low efficiency and high loss, resulting in too slow substrate production capacity increase, which is far from meeting the actual needs of the market. Due to the serious lack of production capacity, the production cost of silicon carbide substrates has remained high. In the cost structure of devices, the substrate accounts for 47% of the cost of silicon carbide devices, which is much higher than the 7% of silicon-based devices.

Please, Log in or Register to view URLs content!
 

european_guy

Junior Member
Registered Member
From what I know, first off, SMIC’s annual report mentioned that SMSC is a consolidated subsidiary of SMIC. Even though they own less than 50%, they have control, so it’s consolidated. SMIC hasn’t hidden SMSC’s revenue, and their overall revenue has been growing fast year-over-year. But, the growth rate could have been higher if it weren’t for the ongoing price drops in the 55nm and lower-end processes over the past couple of years (heard that SMIC’s customers take on the yield losses for their advanced processes). And second, HLMC has managed to develop its N7 node, but the production volume is still quite small.

Thanks for clarification.

Actually in their
Please, Log in or Register to view URLs content!
, when they refer to SMIC, they don't seem to explicitly include SMSC fab. They invite to refer to their website, and there
Please, Log in or Register to view URLs content!
, SMSC seems missing too...

Regarding HLMC, is a big news (at least for me) that they managed to enter trial production with 7nm.

Can I kindly ask if you can share some source with us? Sorry to ask, but this is a big news if confirmed.
 

tphuang

Lieutenant General
Staff member
Super Moderator
VIP Professional
Registered Member
Thanks for clarification.

Actually in their
Please, Log in or Register to view URLs content!
, when they refer to SMIC, they don't seem to explicitly include SMSC fab. They invite to refer to their website, and there
Please, Log in or Register to view URLs content!
, SMSC seems missing too...

Regarding HLMC, is a big news (at least for me) that they managed to enter trial production with 7nm.

Can I kindly ask if you can share some source with us? Sorry to ask, but this is a big news if confirmed.
they used to list SMSC in their half year or yearly report. They stopped breaking down revenue by JVs now. Presumably, to prevent people like me from calculating their wafer production.

There is no reason to believe HLMC is somehow on 7nm.
 

tokenanalyst

Brigadier
Registered Member

JCET Q2 2024 Net Profit Attributable to the Parent Increased by 258% Quarter-on-Quarter, Hitting a Record High for Revenue​

2024-08-23
Q2 2024 Financial Highlights:
  • Revenue was RMB 8.64 billion, an increase of 36.9% year-on-year and 26.3% quarter-on-quarter. A record high Q2 in the company’s history.​
  • Generated RMB 1.65 billion cash from operations. With net capex investments of RMB 0.93 billion, free cash flow for the quarter was RMB 0.72 billion.​
  • Net profit attributable to owners of the parent was RMB 0.48 billion, an increase of 25.5% year-on-year and 258.0% quarter-on-quarter.​
  • Earnings per share was RMB 0.27, as compared to RMB 0.22 in Q2 2023.​
1H 2024 Financial Highlights:
  • Revenue was RMB 15.49 billion, an increase of 27.2% year-on-year.​
  • Generated RMB 3.03 billion cash from operations. With net capex investments of RMB 1.87 billion, free cash flow for the first half of 2024 was RMB 1.16 billion.​
  • Net profit attributable to owners of the parent was RMB 0.62 billion, an increase of 25.0% year-on-year.​
  • Earnings per share was RMB 0.35, as compared to RMB 0.28 in 1H 2023.​

Shanghai, China, August 23, 2024 - Today, JCET Group, a leading global provider of integrated circuit (IC) back-end manufacturing and technology services, announced its financial results for the first half year of 2024. The financial report shows that in the first half of 2024, JCET achieved revenue of RMB 15.49 billion, and net profit attributable to owners of the parent of RMB 0.62 billion, both increased over 25%. In Q2 2024 JCET achieved revenue of RMB 8.64 billion, an increase of 36.9% year-on-year,a record high Q2 in the company’s history, and net profit attributable to owners of the parent of RMB 0.48 billion, an increase of 25.5% year-on-year and 258.0% quarter-on-quarter, generating RMB 1.65 billion cash from operations.
During the reporting period, the company’s capacity utilization has significantly increased, with enlarged investment to expand production capacity of core production lines. In the first half of the year, the revenue of communications, computing, and consumer electronics businesses increased respectively by 48%, 23%, and 33% year-on-year. The company is consistently strengthening the research and development in advanced packaging technologies, with R&D investment of RMB 0.82 billion in 1H 2024, a year-on-year increase of 22.4%.
JCET is also actively promoting strategic projects to enhance its smart manufacturing. After two years construction, the new advanced packaging factory “JCET Microelectronics Wafer-level Microsystems Integration High-end Manufacturing Base” with an area of over 130,000 square meters is progressing with equipment mobilization. The new automotive chip back-end manufacturing base has completed building the factory structure. The acquisition of a high-density memory chip packaging factory has obtained necessary approvals, and the project is progressing towards completion.

Please, Log in or Register to view URLs content!
 

tokenanalyst

Brigadier
Registered Member

Diagnostics of Tin Droplet-Based Laser Produced Plasma by Triple Langmuir Probe​


Huazhong University of Science and Technology

Abstract​

Laser produced plasma extreme ultraviolet light source, as the exposure light source for next-generation lithography, faces urgent challenges due to its low conversion efficiency and significant debris contamination. Using a triple Langmuir probe, we conducted a study on the parameters and kinetic characteristics of tin droplet-based laser produced plasma that serves as extreme ultraviolet light source. In this paper, we report a design of triple Langmuir probe circuit based on the BWL (Bulk Wirewound Low-value) sampling resistor, which ensures that high-frequency signals remain undistorted during acquisition. Utilizing the “shadowgraph” method, we calculated the average deflection angles of droplet debris jets under different alignment conditions. Combining the diagnostic results from the probe, we found that the alignment accuracy between the laser and droplet directly influences the parameters of plasma. The parameters and kinetic characteristics of plasma were also diagnosed and analyzed at different angles and distances by moving the probe. Employing the triple Langmuir probe, measurements and analyses of laser produced plasma parameters and kinetic characteristics were achieved, providing a convenient method for diagnosing extreme ultraviolet lithography light sources. This work also offers an effective basis for optimizing the light sources.

1724459556702.png
1724459609347.png

Please, Log in or Register to view URLs content!
 

tinrobert

Junior Member
Registered Member
I have now switch from writing for Seeking Alpha to writing for SubStack. Here is an article that discusses Applied Materials, sales to China, and WFE equipment growth. The title is

Applied Materials – Loses $750 Million DRAM Business as China’s Chip and Equipment Self-Sufficiency Increase​

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