What's new

Opinionated - China Chipping Away to Semiconductor Dominance

May 8, 2019

Dr. Liang said, "Our FinFET technology R&D is progressing smoothly, as our 12nm is entering customer engagement, and the research and development of our next generation of FinFET is progressing well based on our accumulated technology development. The construction of SMIC South FinFET Fab was completed successfully, and we have begun capacity deployment. We will prepare ourselves to be ready for rapid transitions in customer technology migration to face the ever-changing industry environment."

Did they start mass production of their 16 nm FinFET?
 
now we know why smic is moving so slow. the bosses want to make a tiny profit... those mothafukas.. smic is state-owned company with full government support, why do they even care about profit? they got their priority all wrong. fire em all :D

---------------------------

Co-chiefs of China’s top chipmaker SMIC fighting over strategy
Pair battle over whether to pursue profits or state demands for cutting-edge technology

The co-chief executives of China’s largest chipmaker are fighting over whether to build a profitable business or focus on leading-edge technology, highlighting the problems plaguing the government’s push to build an indigenous semiconductor manufacturing sector.

Zhao Haijun, co-CEO of Semiconductor Manufacturing International Corporation, is considering leaving the company because Beijing policymakers are prioritising the work of co-CEO Liang Mong-song to develop more advanced chip production at 14-nanometre resolution (14 billionths of a metre) and smaller.

This is taking priority over Mr Zhao’s own efforts to expand commercially viable operations using older generations of process technology, according to seven people familiar with the situation. Scores of executives are being replaced to align management more closely with Mr Liang’s goals, the people said.

The upheaval indicates that in China’s current programme for creating a domestic chip manufacturing base, state diktat is gaining the upper hand over commercial objectives.

In 2015, Beijing launched its seventh large-scale plan for developing an integrated circuits industry since the 1950s with the intention of relying on private equity and industry experts, diverging from past approaches that were dominated by state instructions and subsidies.

But analysts said things were slipping back to China’s traditional approach of state instructions and subsidies. “In the second phase of this plan there has been a lot of government money. In China if there’s no government money, people will not join,” said Roger Sheng, an analyst at Gartner. “A lot of government people have been put in charge of these companies recently.”

The fallback to traditional tactics comes as China’s semiconductor manufacturers show no sign of catching up to global leaders in making advanced chips. While SMIC has yet to start 14nm mass production, South Korea’s Samsung was already producing to that standard in 2014.

The problems have been exacerbated by the US government’s attempts to slow China in its pursuit of technological power.

A person close to Mr Zhao said he wanted to leave because the government’s focus on pushing leading-edge technology had made Mr Liang irreplaceable while Mr Zhao was seen as expendable.

An executive who left SMIC recently said entire teams of managers were being pushed out as Mr Liang had been given a free hand to hire people focused on leading-edge technology development. “There is a big purge under way right now,” said a senior executive at a supplier to SMIC. “Zhao and Liang have been fighting for a while, and Liang is winning.”

Mr Liang, a former research and development chief at Taiwan Semiconductor Manufacturing Company, the world’s largest contract chipmaker, joined SMIC in 2017 after a stint at rival Samsung.

Another person with inside knowledge of SMIC said: “The goal under the government plan is that China must achieve 14nm by 2020. Liang is working on that and has actually made progress a lot faster than expected, so the government is fully backing him.”

SMIC did not comment on the situation in detail. But a company spokesperson said: “The rumours are untrue.”

SMIC’s revenue in the first quarter decreased by 19.5 per cent compared with the same period last year to $668.9m, and net profit dropped by 58.2 per cent to $12.3m year-on-year, although the company reversed a fourth-quarter loss. The company reported a gross margin of 18.2 per cent, missing its target of 20-22 per cent.

The company said the gross margin would remain between 18 and 20 per cent for the full year — less than half the level of TSMC.

In a call with investors on Thursday, Mr Zhao said he intended to continue making SMIC’s operations more efficient, cost-effective and competitive, adding that the company had grown strong in speciality chips with “a solid customer base and reasonable demand”.

Analysts said SMIC had managed to gain customers with more mature technology but would struggle to increase profitability.

“They have made a lot of progress on 14nm but it will be very hard to commercialise that because those customers that use 14nm would rather use TSMC,” said an industry expert.


Code:
https://www.ft.com/content/994bca54-7212-11e9-bf5c-6eeb837566c5
 
Last edited:
now we know why smic is moving so slow. the bosses want to make a tiny profit... those mothafukas.. smic is state-owned company with full government support, why do they even care about profit? they got their priority all wrong. fire em all :D

---------------------------

Co-chiefs of China’s top chipmaker SMIC fighting over strategy
Pair battle over whether to pursue profits or state demands for cutting-edge technology

The co-chief executives of China’s largest chipmaker are fighting over whether to build a profitable business or focus on leading-edge technology, highlighting the problems plaguing the government’s push to build an indigenous semiconductor manufacturing sector.

Zhao Haijun, co-CEO of Semiconductor Manufacturing International Corporation, is considering leaving the company because Beijing policymakers are prioritising the work of co-CEO Liang Mong-song to develop more advanced chip production at 14-nanometre resolution (14 billionths of a metre) and smaller.

This is taking priority over Mr Zhao’s own efforts to expand commercially viable operations using older generations of process technology, according to seven people familiar with the situation. Scores of executives are being replaced to align management more closely with Mr Liang’s goals, the people said.

The upheaval indicates that in China’s current programme for creating a domestic chip manufacturing base, state diktat is gaining the upper hand over commercial objectives.

In 2015, Beijing launched its seventh large-scale plan for developing an integrated circuits industry since the 1950s with the intention of relying on private equity and industry experts, diverging from past approaches that were dominated by state instructions and subsidies.

But analysts said things were slipping back to China’s traditional approach of state instructions and subsidies. “In the second phase of this plan there has been a lot of government money. In China if there’s no government money, people will not join,” said Roger Sheng, an analyst at Gartner. “A lot of government people have been put in charge of these companies recently.”

The fallback to traditional tactics comes as China’s semiconductor manufacturers show no sign of catching up to global leaders in making advanced chips. While SMIC has yet to start 14nm mass production, South Korea’s Samsung was already producing to that standard in 2014.

The problems have been exacerbated by the US government’s attempts to slow China in its pursuit of technological power.

A person close to Mr Zhao said he wanted to leave because the government’s focus on pushing leading-edge technology had made Mr Liang irreplaceable while Mr Zhao was seen as expendable.

An executive who left SMIC recently said entire teams of managers were being pushed out as Mr Liang had been given a free hand to hire people focused on leading-edge technology development. “There is a big purge under way right now,” said a senior executive at a supplier to SMIC. “Zhao and Liang have been fighting for a while, and Liang is winning.”

Mr Liang, a former research and development chief at Taiwan Semiconductor Manufacturing Company, the world’s largest contract chipmaker, joined SMIC in 2017 after a stint at rival Samsung.

Another person with inside knowledge of SMIC said: “The goal under the government plan is that China must achieve 14nm by 2020. Liang is working on that and has actually made progress a lot faster than expected, so the government is fully backing him.”

SMIC did not comment on the situation in detail. But a company spokesperson said: “The rumours are untrue.”

SMIC’s revenue in the first quarter decreased by 19.5 per cent compared with the same period last year to $668.9m, and net profit dropped by 58.2 per cent to $12.3m year-on-year, although the company reversed a fourth-quarter loss. The company reported a gross margin of 18.2 per cent, missing its target of 20-22 per cent.

The company said the gross margin would remain between 18 and 20 per cent for the full year — less than half the level of TSMC.

In a call with investors on Thursday, Mr Zhao said he intended to continue making SMIC’s operations more efficient, cost-effective and competitive, adding that the company had grown strong in speciality chips with “a solid customer base and reasonable demand”.

Analysts said SMIC had managed to gain customers with more mature technology but would struggle to increase profitability.

“They have made a lot of progress on 14nm but it will be very hard to commercialise that because those customers that use 14nm would rather use TSMC,” said an industry expert.


Code:
https://www.ft.com/content/994bca54-7212-11e9-bf5c-6eeb837566c5
yes.the most important goal of smic is to catch up in tech. we do not need their tiny profit that is laughable.SMIC need to learn from BOE
 
You will hear more news on SMIC's 7nm process fairly soon.

The days when SMIC talk about 5/3nm process are not far off.

By the way, Huawei's HiSilicon will be a 10 billion dollar fabless semi company by 2020. :D

I have been hearing news of SMIC's 7 nm process for many many years. It is actual results that matter not words. That is why I am more interested in first knowing how their 14 nm process is coming up.
 
Chinese AI start-up launches groundbreaking chip

Source: Xinhua| 2019-05-09 16:50:39|Editor: ZX

SHANGHAI, May 9 (Xinhua) -- During a demonstration on Thursday, an Artificial Intelligence (AI) server the size of a laptop used 200 cameras to complete a real-time intelligent video analysis task -- matching the faces of several hundred audiences members with their previously provided photos with a 100 percent match rate, all within a second.

The AI server, developed by Shanghai-based YITU Technology, has computing power equivalent to eight NVIDIA P4 servers, with half the size and 20 percent of the energy consumption, slashing the construction cost by 50 percent and the operation and maintenance cost by 80 percent for data centers.

YITU's groundbreaking chip "QuestCore," offers a visual analysis performance two to five times faster than the similar products available on the market with the same power consumption.

"Generally speaking, QuestCore is the most intelligent and cost-effective video analysis chip in the world," said Lu Hao, YITU's chief innovation officer, adding that it is 100 percent designed and manufactured in China.

He said the SoC (System-on-Chip), the equivalent of a computer motherboard on a single chip, is conducive to improving China's IC industry and bringing down the cost of AI application and promotion for Chinese companies in fields such as security, healthcare, retail and AI city.

YITU has taken the top place in face recognition tests held by the National Institute of Standards and Technology (NIST) under the U.S. Department of Commerce for three consecutive years.

http://www.xinhuanet.com/english/2019-05/09/c_138045878.htm

https://user.guancha.cn/main/content?id=112887&s=fwzxfbbt
 
I have been hearing news of SMIC's 7 nm process for many many years. It is actual results that matter not words. That is why I am more interested in first knowing how their 14 nm process is coming up.
Go google genius, you think we are working in SMICS. An Indian telling us about future tense? Gosh. For many years? 7nm was only popularized for the past 2 years. hahaha
 
Chinese AI start-up launches groundbreaking chip

Source: Xinhua| 2019-05-09 16:50:39|Editor: ZX

SHANGHAI, May 9 (Xinhua) -- During a demonstration on Thursday, an Artificial Intelligence (AI) server the size of a laptop used 200 cameras to complete a real-time intelligent video analysis task -- matching the faces of several hundred audiences members with their previously provided photos with a 100 percent match rate, all within a second.

The AI server, developed by Shanghai-based YITU Technology, has computing power equivalent to eight NVIDIA P4 servers, with half the size and 20 percent of the energy consumption, slashing the construction cost by 50 percent and the operation and maintenance cost by 80 percent for data centers.

YITU's groundbreaking chip "QuestCore," offers a visual analysis performance two to five times faster than the similar products available on the market with the same power consumption.

"Generally speaking, QuestCore is the most intelligent and cost-effective video analysis chip in the world," said Lu Hao, YITU's chief innovation officer, adding that it is 100 percent designed and manufactured in China.

He said the SoC (System-on-Chip), the equivalent of a computer motherboard on a single chip, is conducive to improving China's IC industry and bringing down the cost of AI application and promotion for Chinese companies in fields such as security, healthcare, retail and AI city.

YITU has taken the top place in face recognition tests held by the National Institute of Standards and Technology (NIST) under the U.S. Department of Commerce for three consecutive years.

http://www.xinhuanet.com/english/2019-05/09/c_138045878.htm

https://user.guancha.cn/main/content?id=112887&s=fwzxfbbt


this is great news. a few days ago hikvision announced that all their nvidia sh1t will be replaced with new generation chips from cambricon. now with this new chip, nvidia is losing another big customer in china. alibaba, baidu and tencent will be next to kick nvidia ***. one by one. lol. great again..

btw, that chip is actually made by chinese startup, thinkforce, not yitu. it's a many-core type cpu for high performance. yityu is thinkforce's biggest investor :D
 
Mr Liang, a former research and development chief at Taiwan Semiconductor Manufacturing Company, the world’s largest contract chipmaker, joined SMIC in 2017 after a stint at rival Samsung.

Another person with inside knowledge of SMIC said: “The goal under the government plan is that China must achieve 14nm by 2020. Liang is working on that and has actually made progress a lot faster than expected, so the government is fully backing him.”

Taiwan is a good case of suckering USers and having easy access to otherwise more time-consuming knowledge.

:D

Come on, suckers, we are democracy :lol:
 
Taiwan is a good case of suckering USers and having easy access to otherwise more time-consuming knowledge.

:D

Come on, suckers, we are democracy :lol:
samsung cought up with tsmc fab technology, because of mr. liang and his team they recruited from tsmc. now he's leading smic and brough along his team with him too. the talk about 5/3nm process are not far off.
 
businesswire.com
Keysight Technologies, Calterah Collaborate to Launch New Generation Automotive Radar Chipset
May 02, 2019 11:00 AM Eastern Daylight Time
5-6 minutes

SANTA ROSA, Calif.--(BUSINESS WIRE)--Keysight Technologies, Inc. (NYSE: KEYS), a leading technology company that helps enterprises, service providers and governments accelerate innovation to connect and secure the world, has joined forces with China's leading automotive millimeter wave radar chipset design house, Calterah, to support the design, verification, test and launch of a new generation of automotive millimeter wave radar chipset that will drive development of the connected car supply chain.

In March 2019 Calterah officially released the ALPS family of chips which integrates high-speed analog-to-digital converters, complete radar signal processing baseband and high-performance CPU core. In the process of design and verification of this chip, Keysight supported Calterah with the company’s E8740A automotive radar signal analysis and generation solution for 77GHz radar signal receiving and transmission characterization, radar target simulation, and testing of the basic radio frequency parameters of the chip. Engineers from both companies worked together to accelerate time to market of the chipsets, while ensuring they deliver the best possible performance.

“Keysight and Calterah have collaborated on the development of automotive radar chipsets for many years, including the first and second generation of chips, and ultimately the automotive industry supply chain,” said Qitao Pan, general manager of the Automotive and Energy Solutions Group at Keysight Technologies China. “Keysight has delivered mature and complete testing solutions for key technologies such as millimeter wave radar, automotive Ethernet, automotive wireless interconnect and V2X. We will continue to develop solutions in cooperation with OEMs, as well chip manufacturers, to support the global automotive industry supply chain.”

“ALPS is a system on a chip (SoC), and its chip architecture and performance parameters are far superior to previous generations,” said Dr. Wang Dian, chief mmW scientist at Calterah. “Keysight's measurement solution covers the entire range from digital to analog to RF, ensuring the functionality and performance of each sub-module of ALPS, as well as the superior performance of the final SoC. We hope to continue our cooperation with Keysight in the future as we collectively support the automotive technological evolution.”

About Calterah

Calterah Semiconductor is a leading provider of 77-GHz CMOS mmWave radar sensor IC targeting ADAS, security screening & imaging, and smart home, etc. It is the only enterprise of China that has mass-produced 77GHz and 79GHz mmWave radar sensor chips so far. In 2017, the company mass-produced the first generation of 77GHz and 79GHz CMOS mmWave radar transceiver IC. The second-generation SoC cascading chip - ALPS - provides customers with radar solutions that are easy to develop, offer improved performance and reduce energy consumption.

About Keysight Technologies

Keysight Technologies, Inc. (NYSE: KEYS) is a leading technology company that helps enterprises, service providers, and governments accelerate innovation to connect and secure the world. Keysight's solutions optimize networks and bring electronic products to market faster and at a lower cost with offerings from design simulation, to prototype validation, to manufacturing test, to optimization in networks and cloud environments. Customers span the worldwide communications ecosystem, aerospace and defense, automotive, energy, semiconductor, and general electronics end markets. Keysight generated revenues of $3.9B in fiscal year 2018.

Code:
https://www.businesswire.com/news/home/20190502005541/en/Keysight-Technologies-Calterah-Collaborate-Launch-New-Generation

---

Ainstein K-77 Long Range Radar featuring Calterah CAL77A2T4R FOWLP Transceiver

A next-generation mid- and long-range, wideband and high-resolution radar sensor for ADAS based on RF CMOS technology using eWLB packaging.

For autonomous driving, Long Range Radar (LRR) is a key technology. For the last few years, devices have been built using SiGe BiCMOS technology, which allows high power and high-resolution radar. But as the number of radars grows as autonomous driving emerges, low power consumption and low-cost chipsets will be needed. Radio frequency (RF) CMOS technology is therefore newly being used in radar chipsets in order to satisfy these requirements. Beside the main players like Texas Instruments (TI) or NXP, small players like Calterah can also compete. Ainstein chose Calterah for its Kanza-77 (K-77) LRR designed for Advanced Driver-Assistance System (ADAS) applications like forward collision warning or assist brake functions, pedestrian and cyclist collision warnings and adaptive cruise control.

The K-77 is built on RF CMOS Technology from Calterah featuring two transmitters and four receivers designed for low power consumption and unit cost. The chipset allows cascading in order to increase the number of receiving and transmitting paths. Using embedded wafer level ball grid array (eWLB) packaging reduces parasitic signals, making this new chipset compact and powerful. Both monolithic microwave integrated circuits (MMICs) are soldered on an asymmetric printed circuit board (PCB) with a hybrid PTFE/FR4 substrate.

The radar is based on 77 GHz technology and provides two measurement modes, mid- and long-range, within a single package. Moreover, Ainstein developed advanced signal processing algorithms. supported by Xilinx’s dual-core ARM® Cortex™-A9 based processing system. It enables detection and tracking of static and dynamic objects up to 64 targets in the vehicle’s path, while allowing to the sensor to be highly customizable for customer applications.

Code:
https://www.systemplus.fr/reverse-costing-reports/ainstein-k-77-long-range-radar-featuring-calterah-cal77a2t4r-fowlp-transceiver/
 
Last edited:
I have been hearing news of SMIC's 7 nm process for many many years. It is actual results that matter not words. That is why I am more interested in first knowing how their 14 nm process is coming up.

SMIC1412.png


Loongson‘s 3A5000/3C5000 processors will be taped out on SMIC’s 12 nm process.

Shanghai Huali Microelectronics' 14nm technology node will be ready for risk production in 2020.
 
Last edited:
View attachment 559459

Loongson‘s 3A5000/3C5000 processors will be taped out on SMIC’s 12 nm process.

Shanghai Huali Microelectronics' 14nm technology node will be ready for risk production in 2020.
if everything goes smoothly that would be the same level as intel as they are still struggling with 10nm process. :D
 
Last edited:

Back
Top Bottom