Chinese semiconductor industry

Robotics: The company's semiconductor equipment business has relatively stable profitability​

Recently, the robot stated in the latest investor relations activity record form disclosed that the company's semiconductor equipment business has a relatively stable profitability, and the relevant core patented technologies are advanced in the country. With the formation of domestic alternative opportunities in recent years Due to the scale effect, it is expected that the comprehensive gross profit margin will increase in the future.

According to his introduction, the company is a high-tech enterprise with robots and intelligent manufacturing equipment as the core, established by the Shenyang Institute of Automation, Chinese Academy of Sciences as the main sponsor. Mobile robot (AGV), special robot" three main products and "industrial robot complete set of equipment, intelligent logistics complete set of equipment, automatic assembly and testing complete set of equipment".

In addition to the "3+3" main business, the company is also laying out high-potential sectors such as the semiconductor field. Since the company undertook the national 02 special project, it began to lay out the semiconductor field, and has formed a series of complete semiconductor equipment represented by clean robots so far. In 2021, the company's semiconductor equipment will achieve revenue of 164 million yuan, accounting for about 5% of the company's total revenue last year.

Huizhou Eagle Electronic Technology is spending 550m RMB to build a PCB plant in Thailand to supply car industries, computers and appliances. Looks like it's quite important to have something like this there to build up supply chain for Chinese companies that are doing their production in Thailand.

Cuda is a potential danger of China GPUs... they are accelerating the development of China's own GPU architecture..

Is this the end of ARM and other legacy architectures? How far is Huawei in RISC-V?

jfcarli said:

Please, Log in or Register to view URLs content!

Is this the end of ARM and other legacy architectures? How far is Huawei in RISC-V?

Click to expand...

Huawei already uses RISC-V chips in some of its business models..

jfcarli said:

Please, Log in or Register to view URLs content!

Is this the end of ARM and other legacy architectures? How far is Huawei in RISC-V?

Click to expand...

I'm not sure why you are asking about Huawei here. The entire Chinese industry is behind RISC-V. There has been several posts about Ventana and China's own effort.

you can read here about the 2nd generation XiangShan, which is probably comparable to Cortex A76 (developed in 2018)

and this is BOSC's comment regarding the release of Ventana. It believes that it's 3rd generation Xiangshan will be 1 year later than Ventana, but China will make it open source.

and I would be very surprised if this cannot be produced by SMIC, since the entire point of going with RISC-V is to be sanction free. 2nd gen Xiangshan is to be produced by SMIC.

I would not put much faith on those numbers by Ventana. The chip has not taped out yet and they have not tested the first silicon. Those performance figures were computed with software simulations of the chip. It would not be the first time some supposed super design turned out to be a dud when when made into actual transistors. Like what happened with Sun Microsystems's Rock CPU.

mossen said:

Please, Log in or Register to view URLs content!

Click to expand...

@WTAN Sir is this the 200W LPP light source that we keep on hearing from the grapevine? IF true then RS laser had a breakthrough, with it maybe a 90W or 120W power source for the mythical 22nm SSA 900 DUVL?

@olalavn Sir we may see a working prototype next year exactly the timeline you gave for IOC in 2024.

mossen said:

Please, Log in or Register to view URLs content!

Click to expand...

Interesting, it seems Huawei and their partners with would need to possess a small scale technology demonstrator to be able to overcome the issue described in the patent. It's not something you would want to rely solely on computational simulation (if that's even possible) and if you were going to patent this process, you would expect the proof of concept to have been verified through some means otherwise the patent is vaporware. Although we can't infer any capabilities or specs like throughput, this is excellent for multiple reasons:

1) Even if this is a small scale prototype/demonstrator, it means they have actively solved or are working on solving many of the workflow sequences for EUV lithography.
2) To be able to solve the experimental problem described in the patent, the prototype/demonstrator must reasonably approximate the conditions encountered during operating scenarios. Therefore, it can be reasonably inferred that the prototype/demonstrator integrates concepts that would be mirrored in a full scale system in terms of key operating principles, sub-systems.
3) The sub-systems and integration into the prototype/demonstrator must be meeting the right levels of precision, accuracy, etc. They may not be there in terms of throughput or other commercial volume specs, but they are sufficient to establish a working proof of concept. There are other components/systems that will need to evaluated for a commercial setup that aren't required in a prototype/demonstrator.
4) The fact that they worked around ASML's patent shows that China is playing its cards smartly. If they wanted, they could just void ASML patents and go with the same implementation but seeing as to how ASML brings alot of value, there is no gain in alienating them when they still bring alot of value as healthy competition (circumstances permitting).

Exciting times, look forward to seeing what else comes forward.