China wants to be a Chip Superpower

17 August, 2016

China formed the High-end Chip Alliance: a group of 27 industry giants, universities and government offices which will lead a coordinated effort to establish an independent leading Semiconductor Industry

China formed the High-end Chip Alliance: a group of 27 industry giants, universities and government offices which will lead a coordinated effort to establish an independent leading Semiconductor Industry

CHINA-SEMICONDUCTORS

The establishment of a local semiconductor industry has been one of the Chinese government’s strategic goals over the last few years. The focused governmental effort begun as early as 2000, but the first attempts have not been successful. Last week, the Chinese efforts have geared up, with the formation of a Chinese High-end Chip Alliance. The alliance includes 27 companies and public organizations, forming a national concert of companies, funding agencies and research institutions.

The official proclamation of the alliance stated that it will focus on the development of processors, memories, sensors, FPGA components, DAC and ADC signal convertors and various other essential components for the establishment of a fully independent semiconductor industry.

The founding 27 members of this alliance include Tsinghua Unigroup, Yangtze River Storage Technology, Huawei, SMIC, ZTE and China Academy of Telecommunication Research (a branch of the country’s Ministry of Industry and Information Technology). Loongson Technology will coordinate the development of powerful processors for the alliance, while Tsinghua Unigroup, established by the Chinese government, will help finance the founding of a strong local semiconductor industry.

“This alliance of government, academia and industry aims to create a complete ecosystem for domestic semiconductor manufacturers”, said Jian-Hong Lin, research manager of TrendForce. “If successful, the alliance will create a chip industry chain starting from chip architecture to chip production, operation systems, devices, platforms and finally to the IT service market. In sum, this move is another indication of China’s ambition to transform itself from a major manufacturing country by export volume to a global manufacturing leader in terms of product quality.

A processor manufactured by Chinese Loonson
A processor manufactured by Chinese Loonson

In a blog published in the company’s site, Lin explained that the Chinese move aims to elevate china from a mere producer of chips to a global industry leader. ““Taiwanese semiconductor companies cannot survive on just the demand from the domestic market and compatriot electronics brands. This is especially true for the local IC design houses. Their long-term growth will depend discovering new sources of demand and application needs in the international market. Still, China currently is the largest market and has the largest client base for Taiwanese IC design houses. Whether Taiwanese IC industry is allowed to form effective joint ventures or strategic partnerships with the Chinese counterpart is an issue that Taiwan’s government and technology enterprises need to address after the establishment of the high-end chip alliance in China.”

An efficient management of such a national alliance is not an easy task. For example, Tsinghua Unigroup subsidiary Spreadturn develops 16 nanometer chips, while Chinese chip production contractor SMIC has not yet acquired 16 nanometer geometry production capabilities. Nevertheless – the parties joined in the new alliance have shown in the past, that a continuous effort can help them to achieve impressive feats.

Loongson Technologies for example, demonstrates how a Chinese company can achieve a goal which only recently would have been perceived as impossible. The company was founded as part of the Chinese effort to develop a processor based on independent architecture for supercomputers. The company began its life as a project of the government owned ICT. The company has developed a uniquely robust CPU architecture of a 64 core processor, which is intended to grow to a thousand core one.

American mistake which built Chinese the industry

Loongson’s success is the result of an American mistake: during the nineties, Digital (acquired by Compaq, later acquired by HP) has developed a RISC architecture dubbed Alpha, which is still considered one of the best processor families in industry, providing 64 bit processing capabilities as soon as 1991. All this did not save the brand, which was “killed” by Digital and Compaq due to short term business considerations. The companies allowed Chinese acquire the unique technology.

Inside Smic's Fab. source: ARM Connected Community
Inside Smic’s Fab. source: ARM Connected Community

The Chinese chose this discarded architecture, due to the understanding that it saves them the enormous challenges of developing a new architecture fro scratch. In April 2012 they have been proven right: The ICT institute and Loongson Technologies produced a processor in the already than obsolete 65 nanometer geometry. Nevertheless, the processor had a calculation power matching these of concurrent Intel processors. In fact, the same processor was integrated into the Chinese inan Petaflop Shenwei supercomputer, which is considered more efficient in its use of energy than American counterparts.

Loongson’s processor was dubbed Godson – a worth remembering considering that Loongson will coordinate Chinese processor architecture development, and this will be based on the Godson processor.

Tsinghua’s ambitious goal

Another name worth remembering is that of major financing house Tsinghua Unigroup. In 2013 the chines government decided to make a long term investment in the local semiconductor industry. After a few failed attempts as early as 2000, the Chinese government tried a new strategy: instead of trying to develop new technologies, it would establish an independent investment fund which will pursue a strategy aimed to encourage local companies to merge in order to form semiconductor giants.

In September 2014, Intel has made a massive investment in Tsinghua Unigroup, acquiring 20% of its shares for some $1.5 billion. Intel said that the investment’s goal was to expand the use of Intel’s computing platform in the Chinese mobile market. The investment agreement has two very interesting features: Intel is interested in conducting Architecture research and joint projects in China. As well as to develop mobile communication solutions.

In July 2015, Tsinghua tried to acquire American memory components producer Micron, offering a staggering $23 billion. Micron is considered the world’s second largest memory component producer, and its sale to a Chinese group did not go through – probably due to active discomfort by US administration.

The failure of the attempt to take control of Micron does not seem to deter Tsinghua Unigroup from making further strategic investment outside china. In September 2015, Tingsghua Unigroup chairman Zhao Weigo told Reuters of his plans to invest $47 billion over the next five years to to build the world’s third-biggest chipmaker. No wonder the Taiwanese are feeling uneasy.

Share via Whatsapp

Posted in: Business , Electronic Components , Electronics Industry , Featured Stories , News , Semiconductors , Technology