The early 2000’s saw the birth of a groupof Chinese fabless design firms whose experiences paved the way for China’scurrent semiconductor industry. Among these were two designers of mobile chipsets – HiSilicon and Spreadtrum – whose beginnings and growth followed vastly different trajectories, despite their related fields of expertise.
HiSilicon: A Huawei-Supported Design Powerhouse
HiSilicon was formerly Huawei’s application-specific IC (ASIC) design group before it was spun out asan independent entity in 2004. The company is still wholly owned by Huawei, and continues to serve as the telecommunications giant’s chip design arm.
HiSilicon’s close ties with its parent company, which was founded in 1991 and by 2004 had found a firm footing with 46 billion RMB of annual revenue, endowed the new subsidiary with a much stronger technological foundation (and deeper pockets) than the vast majority of other companies at its young age. Supported by this safe foundation, HiSilicon took its time exploring various product opportunities beyond its services to Huawei. The company’s first attempt at a SIM card chip product hit a wall due to plummeting market prices, but it eventually hit its stride in 2007 making chips for security cameras, followed by routers and smart phones. In 2014, HiSilicon launched its flagship Kirin smartphone chipset. Its strong R&D capability shone through as benchmark tests of Kirin showed comparable or better performance to leading edge chips from international leader Qualcomm.
HiSilicon has thus developed five main chip lines - Kirin for smartphones and tablets, Kunpeng CPU’s for data servers, Ascend for AI applications, and Tiangang and Balong for 5G devices and base stations – in addition to other specialized chips for TV’s, cameras and IoT. Since the start, Huawei has continuously served as HiSilicon’s main customer, accounting for up to 90% of the design firm’s sales, and depended on chip technology to differentiate its smartphones, servers, and base stations in from competitor products. Today, HiSilicon has grown into the largest mainland Chinese IC design firm with over 7000 employees, and dominates the global patent landscape for 5G technology.
Number of 5G patents held by various companies
In Q1 2020,HiSilicon became the first mainland Chinese company to break into the top 10chip design firms in terms of global sales. However, this was far from a celebratory moment. HiSilicon’s explosive revenue growth in 2020 Q1 was fueled by Huawei’s stockpiling chips in anticipation of supply shortages due to US sanctions. From 2018 to 2021, alleged espionage and security risks around Huawei’s 5G equipment spurred US regulators to impose export controls in order to cut off both Huawei’s supply of chips and HiSilicon’s access to advanced chipmaking capacity. Unable to manufacture the chips it has designed, HiSilicon was forced to cease selling its signature Kirin chipsets as of September 2020. Skepticism continues to mount around the future of the company. The dramatic rise and fall of China’s premier chip design firm embody the reward and risk of waltzing so closely with a giant. Huawei’s growth has been meteoric, but the company now viewed as a threat to the digital future of the Western world. As Huawei’s closest partner and dependent, HiSilicon must too bear the consequences.
Spreadtrum:The Retun of Overseas Talents
Spreadtrum was established in 2001 by graduates of Tsinghua University, China’s top engineering school, who had studied and worked in the US before returning to their birth country. These founders sensed an opportunity in China’s rapidly marketizing but still immature semiconductor industry, and led a team of over 30 overseas experts to Shanghai’s Zhangjiang Hi-Tech Park (where SMIC happened to be building its first fab), to start designing communications chips for the budding mobile phone market.
Spreadtrum’s founding team had cut their teeth in top Silicon Valley companies, and encountered many shock transitionings to China’s fledgling market economy. To start, while HiSilicon had the devoted support of a well-established firm, independent startups in those days had few options for funding. The venture capital (VC) industry had yet to develop in China. Eventually, it was cross-Pacific investment from Silicon Valley VC’s that enabled Spreadtrum’s beginning, banking on the promise that the team would design chips for China’s next-gen telecommunications standard, 3G.
Other challenges quickly emerged. Bigmobile phone brands were hesitant to use chips from a small new Chinese designfirm. Spreadtrum’s main market would have to be the emerging field of 山寨 shanzhai (literally “mountain village”) phones - off-brand, sometimes copycat, mobile phones catering to the low-end market - made by a smattering of similarly small phone design companies. These customers had little technical capacity to build the systems around a chip, and Spreadtrum soon found itself having to provide the full solution – from the internal mobile hardware to the phone operating system – in order to sell its chips, and still having to maintain a cost palatable to the low-end mobile market.
By 2007, Spreadtrum had not yet scaled itsoriginally-planned 3G telecommunications solution, due to a delay in the determination of China’s national 3G standard. However, the company had succeeded in selling millions of previous-gen (2G) chips to shanzhai phone makers,and eventually some name brands like Samsung. This allowed Spreadtrum to go public on NASDAQ in 2007 to raise an additional $100 million (approx.). In 2013, it was acquired by Chinese SOE Tsinghua Unigroup for $1.78 billion and merged with another IC design firm RDA Microelectronics to form Unisoc (紫光展锐). However, Unisoc’s performance since the merger has been lack luster. Although it unveiled its first 4G LTE chip in 2016, Unisoc failed to win major phone-maker customers from international leaders like Qualcomm and Taiwan’s MediaTek.
To add insult to injury, Unisoc’s parent company Tsinghua Unigroup defaulted on a $200 millionbond in November 2020, triggering scrutiny that unveiled a precarious debt position. Tsinghua Unigroup’s fate is still being decided, with involvement from government decision makers, financial institutions, and the university that remains its majority shareholder. However, some see an opportunity in Unisoc’s organizational uncertainty and larger tension of the China-US technology competition – as the dust settles from the struggles of Tsinghua Unigroup and HiSilicon, perhaps there will arise renewed room for companies like Unisoc to drive the future of China’s semiconductor industry.
The next chapters to HiSilicon and Unisoc’s narratives are still being written, but theirhistories embody the unique factors faced by design firms in China’s rapidly transforming semiconductor industry.
All opinions expressed in this essay represent my personal views only.