Chinese semiconductor thread II

tokenanalyst

Brigadier
Registered Member

Tsinghua University research team makes new progress in high-frequency supercapacitor research​

With the rapid development of artificial intelligence and high-performance computing, computing power and power demand are growing exponentially, which poses a dual challenge to the power supply density and efficiency of power management chips. In this context, power management chips are moving towards on-chip integration of passive components to achieve high-density three-dimensional power supply. However, the performance density of traditional silicon-based passive components has approached the physical limit and is difficult to meet the demand.
Dr. Gordon Moore, founder of Intel and the originator of Moore's Law, pointed out (Proceedings of the IEEE, 1998, 82): The lack of large-capacitance capacitors and inductors is the fundamental bottleneck for the development of integrated electronics. Although the micro-electrochemical supercapacitors developed in recent years have shown high capacitance density characteristics, their inherent static characteristics make it difficult to apply to integrated circuits that mainly use AC high-frequency signals.
Recently, Wang Xiaohong's team from the School of Integrated Circuits at Tsinghua University made a breakthrough in the research on the dynamic response limit of high-frequency supercapacitors. This research quantitatively measured the upper limit of the dynamic response frequency of supercapacitors through experiments for the first time.
The research team used micro-nano processing technology to construct an absolutely flat ideal electrode with a pore-free structure, and eliminated interference through methods such as parasitic capacitance shielding layer structure and external phase-locked loop amplification, thereby accurately determining the upper limit of the dynamic response frequency of the supercapacitor for the first time.
On this basis, the team innovatively proposed the concept of "dielectric-electrochemical" asymmetric capacitors - the device is mainly based on electrochemical effects in the low frequency band and dielectric effects in the high frequency band, achieving a double breakthrough in frequency response and capacitance density. The characteristic frequency of the micro-supercapacitor chip prepared based on this concept exceeds 1MHz, which is six orders of magnitude higher than commercial supercapacitors and covers the operating frequency band of mainstream power circuits.​

Please, Log in or Register to view URLs content!
 

tphuang

General
Staff member
Super Moderator
VIP Professional
Registered Member
SMIC reportedly scaled up from producing just over 20,000 12-inch wafers per month to over 30,000 wafers per month by early 2025, which is a 50% jump. That means Cambricon’s 7nm AI processors, once manufactured by TSMC, can now be produced domestically by SMIC.
Please, Log in or Register to view URLs content!

I don't think this has been posted yet. From people talking to Taiwan's semi industry. I think the 100k wpm of 7nm capacity part is the most interesting part.

A senior executive from a Taiwanese semiconductor firm recently attended several investment forums held by Chinese funds and was surprised by the overwhelmingly upbeat sentiment he encountered. The consensus view was that SMIC’s 7nm capacity could surge to over 100,000 wafers per month by 2026, putting it in the same league as TSMC.

“They’re quite confident about breaking into 5nm,” the executive said. In China, the push toward 5nm has been elevated to a national-level initiative, complete with dedicated task forces coordinating across domestic equipment, materials, and subsystems suppliers. “Right now, it’s full steam ahead.”
 

tokenanalyst

Brigadier
Registered Member

Total investment of 5 billion yuan! Lead Technology Integrated Circuit Components Base Project landed in Shanghai Lingang​


Shanghai Lingang issued a statement saying that on the morning of May 22, the Lingang New Area Management Committee, Lingang Group, and Pioneer Technology Group jointly signed a project investment agreement in Guangzhou, and the Pioneer Technology Integrated Circuit Core Components and System Headquarters Base Project was officially launched in the Lingang New Area.

According to reports, the project will invest 5 billion yuan and use 108 acres of land to build an integrated circuit core parts base in the Lingang New Area, including mass flow controllers, semiconductor core metals, vacuum systems, lithography systems, and standard production lines such as micro-nano processing, defect detection, and online detection. The implementation of the project not only marks the gradual transformation of Pioneer Technology Group from a leader in rare metals to a platform-based semiconductor core parts enterprise, but also has important significance for the development of Shanghai Lingang's integrated circuit core parts industry. From the initial contact to the signing on May 22, the project took only two months to land, which once again interpreted the "Lingang speed" and also reflected the confidence and determination of the Lingang New Area to firmly develop the hard-core manufacturing industry.

According to data, Pioneer Technology Group was founded in 1995. It is a national high-tech enterprise group that originated from the rare metal business and is a leading semiconductor high-end materials, devices, modules and system R&D, production, sales and recycling services. It has two listed companies. In 2024, the group's revenue will reach 40 billion, with more than 10,000 employees and outlets in 17 countries around the world.

Please, Log in or Register to view URLs content!
 

huemens

Junior Member
Registered Member
More info about the new China-specific Nvidia chip.

Please, Log in or Register to view URLs content!

Exclusive: Nvidia to launch cheaper Blackwell AI chip for China after US export curbs, sources say​

The GPU or graphics processing unit will be part of Nvidia's latest generation Blackwell-architecture AI processors and is expected to be priced between $6,500 and $8,000, well below the $10,000-$12,000 the H20 sold for, according to two of the sources.
It will be based on Nvidia's RTX Pro 6000D, a server-class graphics processor and will use conventional GDDR7 memory instead of more advanced high bandwidth memory, the two sources said.
They added it would not use Taiwan Semiconductor Manufacturing Co's advanced Chip-on-Wafer-on-Substrate (CoWoS) packaging technology.
The three sources Reuters spoke to for this article declined to be identified as they were not authorised to speak to media.
Chinese brokerage GF Securities said in a note published on Tuesday that the new GPU would likely be called the 6000D or the B40, though it did not disclose pricing or cite sources for the information.
According to two of the sources, Nvidia is also developing another Blackwell-architecture chip for China that is set to begin production as early as September. Reuters was not immediately able to confirm specifications of that variant.
Investment bank Jefferies estimates that the new regulations cap memory bandwidth at 1.7-1.8 terabytes per second. That compares with the 4 terabytes per second that the H20 is capable of.
GF Securities forecast the new GPU will achieve approximately 1.7 terabytes per second using GDDR7 memory technology, just within the export control limits.
 

tokenanalyst

Brigadier
Registered Member

Nankai University: Optically readout millimeter-wave focal plane array chip based on "metamaterials + MEMS"​


Recently, a research team led by Associate Professor Gong Cheng and Professor Liu Weiwei from the Institute of Modern Optics at Nankai University invented an inexpensive MEMS focal plane array stacked chip that can use light to achieve high-sensitivity millimeter wave detection and high-speed imaging. The chip has 1,600 detection units, with a measured frame rate of 144 Hz, and can detect transient structural morphological changes in matter. This technology embodies the great advantages of optoelectronic fusion and is a new attempt in the field of microwave photonics. The research results were published online in " Light: Science & Applications " under the title "Using light to image millimeter wave based on stacked meta-MEMS chip". Doctoral student Wang Han is the first author of this article, Associate Professor Gong Cheng of Nankai University is the corresponding author, and Nankai University is the first completing unit. The collaborative research team of this article is the team of Professor Yan Bo and Researcher Wang Zhigang from the National Key Discipline of Electromagnetic Field and Microwave Technology of the University of Electronic Science and Technology of China.

1748100693818.png

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