828cloud

Data, Info and News of Life and Economy

Tag Archives: Semiconductor

The Rising Tide of Semiconductor Cost


See large image . . . . . .

Doug wrote . . . . . . . . .

There’s a quiet upheaval happening in the semiconductor industry. The rules that have always governed the industry are fraying, undoing assumptions that we took for granted, that was pounded into us in school. The irreproachable Moore’s Law, that exponential progress will make things cheaper, better, and faster over time, is dead.

People are starting to appreciate that making a chip is not easy. Shortages and the geopolitical concentration of TSMC and ASML have awakened the popular imagination and have highlighted the science-fiction-like process of chipmaking. The road ahead has obstacles that aren’t widely appreciated. Making a semiconductor is going to get even harder, more expensive, and more technical. In other words, the challenges are going to accelerate.

To operate in the future, chipmakers will need more scale, more talent, and more money. I’ve written about this before, but I want to dive deeper into what’s driving the rising costs of making a semiconductor. It affects the entire range of chips, from the most advanced chips to the most basic. The trend is not new, it’s already been happening, but I believe now it will start to pick up speed. And the price increases will likely impact every person on earth. This inflationary cost is not transitory.

To understand how we got here, I want to first refresh you on the death of Moore’s Law. We’ve topped out the growth in transistor-energy scaling, frequency scaling, and we’re starting to hit the end of multi-core scaling in transistor-density increases. But more important than the end of those trends, cost scaling has ended. While we continue to improve transistor density through new techniques, each one layers additional costs.

ASML, in its investor day, made a bold statement that Moore’s Law will continue with system-level scaling. Another name for this is advanced packaging. But these costs are additive to the already escalating costs of making a smaller transistor.

While I believe Advanced Packaging is going to solve the transistor-density problem, I don’t believe it will make chips cheaper. In fact, transistors per dollar have gotten more expensive since the early 2010s.

I want to focus not just on the technological headwinds, but the cost headwinds. One of the major historical assumptions of Moore’s Law is that not only would your transistors double every two years, but the cost of the transistors would decline. No longer. The chart below is from Marvell’s 2020 investor day. The bar for 28nm was approximately 2011-2012.

What’s interesting is that a qualitative change happened around 28nm, as it was one of the last planar nodes. Planar in plain language is a two-dimensional surface (plane), while FinFET – the technology that replaced planar – introduced a “fin” into the transistor to jut upwards, creating a 3D structure instead of a 2D structure. We are now on the verge of yet another gate transition – gate-all-around (GAA), which is an even more 3D-intensive structure. As we switch to GAA or the next iteration of gate technology, I believe that the cost increases per 100m gates will continue to increase, just like they did for FinFET over planar. This is driven by the increased complexity of making these chips — namely, the added number of steps in manufacturing.

It isn’t just this transition that’s pushing costs higher. The lagging edge — older chips — is starting to get more expensive, too. The story here is not technological, but rather economic, and what was once ample capacity with commodity-like returns is starting to become in-demand. Businesses are not willing to add capacity unless subsequent price increases follow. This is another key driver, not just on the most advanced but in the older chips as well.

Finally, not just old chips and new chips, but the companies that make the chips (semiconductor fabs) are becoming more consolidated and more strategic. There really isn’t a lot of room at the leading edge, where the most advanced chips are made. This is the third driver of semiconductor costs, that fabs that offer a one-of-a-kind product are passing their rising costs on to customers. TSMC is not a price-taker, and the world is reliant on its products. Despite the increasing costs, they’re are starting to extract larger profits. Fabless companies have no choice but to pay more.

Each of these themes deserves a deeper dive. I’ll start first with my favorite topic: Semicap – or the tools that are required to make a chip.

Industry Consensus: Semicap Cost Intensity Will Go Up

One of the universal themes this earnings season was the higher cost of tools to make semiconductors.

The drastic price increase is broad-based and is across DRAM, NAND, and Logic. Given that 5nm is in production today, this is not a prediction, but a trend that’s going to continue. The primary driver is not only the rising costs of tools such as EUV, but the rising number of steps to make a chip. Below is a graphic that shows the increase of steps over time.

It’s just not Tokyo Electron making the call for higher intensity alone. This most recent earnings season, TSMC, Lam Research, KLAC, and other Semicap companies called out rising intensity. I think all else being equal, the cost of 100K wafer starts should start to rise low- to mid-single digits per chip at the leading edge. Another way to look at this is from the top-down perspective. I compared the total capacity shipped in Million Square Inches (MSI) and compared this to wafer-fab equipment growth. Think of this as total volume versus the spending to make more wafers.

[ . . . . . . . . . ]

Read more at Fabricated Knowledge . . . . .

Taiwan’s TSMC to Build First Chip Plant in Japan

Yuri Kageyama wrote . . . . . . . . .

Japanese electronics maker Sony and TSMC of Taiwan said Tuesday they plan to jointly build a computer chip plant in Japan with an initial investment of $7 billion.

The plant in the southwestern city of Kumamoto will be the first foundry in Japan for TSMC, one of the world’s leading chipmakers. Construction will start next year for the plant to be up and running by 2024, employing 1,500 high-tech professionals.

The move comes as a supply crunch in chips has slammed various Japanese companies, including automaker Toyota Motor Corp. and video-game maker Nintendo Co., as lockdowns and other coronavirus measures in parts of Asia have hurt chip production.

Sony Semiconductor Solutions Corp., a wholly owned subsidiary of Sony Corp., will be a minority stakeholder, with less than 20% equity, investing $500 million in the new TSMC subsidiary, Japan Advanced Semiconductor Manufacturing.

The overall investment is getting “strong support from the Japanese government,” the joint statement said.

The deal is subject to regulatory approval.

Terushi Shimizu, president of Sony Semiconductor Solutions, said the global semiconductor shortage was expected to continue. The partnership with TSMC will help not only Sony but also other companies, he said.

There has been speculation other Japanese companies, such as Toyota Motor Corp., may join the project, but TSMC declined comment, saying nothing was decided.

“We are pleased to have the support of a leading player and our long-time customer, Sony, to supply the market with an all-new fab in Japan, and also are excited at the opportunity to bring more Japanese talent into TSMC’s global family,” said Chief Executive C.C. Wei.


Source : AP

北京大学成立集成电路学院

记者: 孙竞 . . . . . . . . .

日前,北京大学宣布成立集成电路学院。记者了解到,北京大学集成电路学院以“多学科交叉、产教融合、注重实践”为办学特色,着力集成微纳电子、电子设计自动化、集成电路设计、集成电路制造、集成微纳系统五个重点方向,培养交叉复合型集成电路人才,提高集成电路科学研究与工程应用水平,与国内龙头企业携手支撑和推动集成电路技术与产业的持续发展。

集成电路学院名誉院长王阳元院士在成立仪式上指出,当前“天时、地利、人和”的条件是我国集成电路发展的最佳时期,集成电路已进入了高速发展期。自力更生、掌握核心技术、解决“卡脖子”问题已成为“从庙堂之高”到“江湖之滨”广泛的共识。他认为,学院头等大事是抓好人才培养工作,要培养出爱国、奉献、“敢将热血洒春秋”的高素质人才;同时要坚持原始创新,加强基础和应用基础研究工作,并且将我们的原始创新及时地转化到产业中去。

据悉,北京大学集成电路学院将加强与北京大学计算机、数学、物理、化学、材料等多个优势学科的交叉融合,深化与集成电路产业多环节龙头企业的合作,构建“人才培养、科学研究、产业促进”三位一体的集成电路创新生态,建设具有北大特色的“集成电路科学与工程”一级学科,打造国际一流的集成电路人才培养和科技创新高地。

记者观察到,截至目前,国内已有多所高校成立集成电路学院。7月14日,华中科技大学集成电路学院正式揭牌。该学院以服务国家重大战略和区域经济发展为目标,建设存储器、传感器、光电芯片、显示器、化合物半导体等特色方向。4月22日,清华大学成立集成电路学院,在国内首次提出“1+N”联合机制,布局纳电子科学、集成电路设计方法学与EDA、集成电路设计与应用、集成电路器件与制造工艺、MEMS与微系统、封装与系统集成、集成电路专用装备、集成电路专用材料等研究方向。


Source : 人民网


Read also at Peking University

芯怀天下 集成梦想——北京大学成立集成电路学院 . . . . .

紫光被申破產重整 內地半導體強國夢再受挫

作者: 薛偉傑 . . . . . .

7月9日,內地的紫光集團發表公告透露,在當日收到北京市第一中級人民法院的通知,其債權人徽商銀行以紫光集團不能清償到期債務、資產不足以清償全部債務兼且明顯缺乏清償能力、具備重整價值和重整可行性為由,已向法院申請對紫光集團進行破產重整。該集團將依法全面配合法院進行司法審查,積極推進債務風險化解工作,支持法院依法維護債權人的合法權益。

紫光集團又回應內地媒體表示,被債權人申請破產重整並未對集團屬下公司的日常生產經營造成直接影響,目前集團屬下公司生產經營活動均正常開展。而法院是否受理、集團是否進入重整程序,尚存在不確定性。

早在大半年前,紫光集團已被踢爆出現周轉困難,到去年底已無法償還多批到期的境內外債券,出現多宗債務違約事件,當時該集團已強調,積極推進債務風險化解工作。

但大半年後,仍然發展至被申請破產重整,不能不說是令人失望。事件曝光後,成為兩岸網絡上的大新聞,更被認為是內地半導體強國夢的一大挫折。

曾豪言「併購聯發科、收購台積電」

由2013年起,紫光集團通過多次併購,成為了內地最大的綜合性半導體企業。2015年,紫光集團董事長趙偉國到訪台灣,更豪言要「併購聯發科、收購台積電」,其財大氣粗的作風令人側目。趙偉國在內地媒體面前也毫不避諱他鍾情資本運作方式,但目前紫光集團正面臨被自己掀起的資本風暴吞噬的危機。

在趙偉國的主導下,由2013年開始,紫光集團就開始了「併購之旅」:先後收購在美國上市的晶片設計公司展訊通信、物聯網晶片公司銳迪科微電子,HP惠普旗下的雲端網路設備公司華三通信。之後,還斥資38億美元取得硬碟生產商Western Digital的15%股權,成為最大股東。還收購法國微連接器公司Linxens接近100%股權。甚至曾經打算收購美國最大的記憶體生產商Micron,只是美方不接受。在大約6年時間,紫光集團以及其屬下企業先後對20多家企業發起了併購要約,投入資金超過1000億元人民幣。

「短貸長投」造成財務危機

有些觀點認為,由於起步較遲,紫光集團主要通過大規模併購,來變成半導體巨頭,屬無可厚非,而該集團亦已經擁有「國際級競爭力」。但亦有人擔憂,半導體行業資金技術非常密集,投資回報周期長,倚靠高槓桿的融資方式來收購,會令集團的負債規模過大,財務結構失衡,暗藏風險。現在看來,顯然是後一種觀點較為準確。

問題爆發的確源於融資結構失衡。因為紫光集團的負債較多為高息的短債,但投資回報卻是相當長的事情。加上當初的估計過份樂觀,旗下有部分公司遲遲未能獲得顯著的市場佔有率,卻需要持續投入資金,例如生產記憶體晶片的「長江存儲」,以及設計手機處理器的紫光展銳等。於是,該集團的流動資金嚴重吃緊,要靠不斷發新債來償還舊債或應付營運所需。

據內媒報導,截至2020年6月,紫光集團的總負債規模達到2029.38億元人民幣,比2012年底的46.47億元人民幣暴升了接近44倍。更值得關注的是,在這2029.38億元的債務當中,流動負債為1192.11億元。此外,融資環境也發生了變化。在2019年第一季之後,紫光集團就開始難以通過債券市場持續發新債券,因而在去年底連環爆出債務違約事件。

一般認為,在趙偉國2009年以他的健坤集團購入紫光集團的49%股份之前,紫光集團的情況並不算好。他入主紫光集團初期,的確有拯救和壯大該集團的功勞。而他主導紫光集團以160億元人民幣的高價收購展訊和銳迪科兩間晶片設計公司之後,瞬即獲Intel出資90億元人民幣購入紫光集團旗下併購展訊和銳迪科的主體公司紫光展銳的20%股權,以公司價值計,的確是「有賺」。對此,趙偉國曾得意地說:「企業併購是花錢的,但紫光通過併購發展壯大,我們把收購變成了賺錢的方式。」

但問題是,趙偉國有些貪勝不知輸,過份相信自己的併購和高槓桿營運能力,最終引致資金鏈斷裂。有些財資界人士指出,內地的民風過於急功近利,只求在最短時間內快速進入行業,利用高資本槓桿來買到規模及地位,而非發展壯大。而有些內地企業亦覺得,只要跟著政府政策發展,自己就一定不會是輸家,出了事也會有人埋單。

無疑,以紫光集團現時的規模和重要性,顯然是「大到不能倒下」。而它旗下的幾間晶片設計公司以及生產記憶體晶片的長江存儲,也是比較值錢的。此外,紫光集團還持有中芯國際的股份,是後者的主要股東。但為了解決債務問題,它很可能要賣出這些比較值錢的公司的部分股權。

最重要的是,這次事件對於內地半導體企業日後的融資、收購合併、吸引境外人才(包括台灣人和美籍華人)等等,肯定會帶來不良影響。內地官方和企業都必須吸取其教訓。


Source : Ming Pao

Three Months, 700 Steps: Why It Takes So Long to Produce a Computer Chip

Jeanne Whalen wrote . . . . . . . . .

Christopher Belfi was waiting tables in a lakeside resort near this Upstate New York town a decade ago when he got the career break he’d been waiting for — an invitation to work at a semiconductor factory.

Belfi, who’d recently graduated from the State University of New York at Albany with a technology degree, started chatting with two customers who turned out to be managers at the nearby semiconductor factory. “I used to coach robotics teams in college. And so we were just talking about that. They left their business cards in my checkbook, and I applied and never turned back,” Belfi says.

He started out as a technician, repairing the automated equipment that carries silicon wafers from machine to machine. In time, he rose to oversee the automated system that propels thousands of pods along tracks on the ceiling, each carrying 25 shiny discs that will someday power a mobile phone, an airplane or an automotive air bag.

Lately there are more pods than ever crowding those tracks, as the GlobalFoundries plant — one of a handful of similar factories in the United States — races to keep up with soaring demand for computer chips.

The tiny components are the brains behind an ever-growing array of electronics, from toothbrushes and refrigerators to vacuum cleaners and cars. Global chip sales are forecast to grow by 20 percent this year and by 9 percent next year as smartphones and laptops use more of the components. Even the most mundane products — tires, doorbells and lightbulbs — now require chips to make them work.

Yet the huge expense of building a semiconductor factory — and the months-long process needed to make a chip — means global demand is far outstripping supply. That has forced automakers and other chip users to idle production and prompted lawmakers to endorse federal subsidies to try to boost U.S. chip manufacturing.

Countries lavish subsidies and perks on semiconductor manufacturers as a global chip war heats up

There are hundreds of companies that design computer chips but fewer than two dozen globally that manufacture them in large quantities, leaving those factories under huge pressure.

Many of the biggest facilities are in Taiwan, which now produces 20 percent of the world’s semiconductors and over 90 percent of the highest-tech chips, according to a report commissioned by the Semiconductor Industry Association, which calls itself the “voice of semiconductor industry.”

Asia as a whole produces about three-quarters of global semiconductors, while the United States manufactures about 13 percent. To boost U.S. production, the Senate authorized $52 billion in subsidies last month for new factories and chip research. The measure, supported by President Biden, still must clear the House, where it has yet to be added to pending legislation.

GlobalFoundries, which is wholly owned by the government of Abu Dhabi and has its headquarters in Malta, 30 minutes north of Albany, is one of the main sources of U.S. production. The Malta plant runs 24 hours a day, pumping out 500,000 intricate silicon wafers a year that are then cut into individual chips.

Hundreds of employees don elaborate protective gear known as bunny suits at the start of each shift, to prevent stray lint or hair from marring the wafers. Even a speck of dust can ruin the painstaking process for producing the chips.

Before giving The Washington Post a tour on a recent morning, Belfi suited up: two layers of booties, two pairs of gloves, a hairnet, a hood and a jumpsuit. Protective goggles and face masks are also standard, so adapting to covid-19 protocols “really wasn’t any different,” Belfi said.

He manages 95 people and has learned to identify them in their suits. “You recognize people by the stride they walk at,” he said.

The factory is filled with the static hum of $10 billion worth of machinery and the warm glow of yellow light, which protects the light-sensitive wafers from damage.

The 12-inch silicon discs look like shiny, smooth mirrors when they arrive at the factory. Three months later they are covered with intricate etchings forming billions of transistors, the microscopic switches that control electric currents and allow the chip to perform tasks.

There are about 700 processing steps along the way, through which dozens of layers of patterns are printed and etched on top of each other, following designs provided by each chip customer.

“Think about making a cake,” Belfi said over the noise of the machinery. “In this case it’s going to be a 60- to 75-layer cake, and that cake is built over approximately two and a half, three months.”

To create each layer, the wafer is coated with a light-sensitive chemical. Then a high-tech printer known as a lithography machine projects the same tiny pattern over and over across the wafer, as if it were stamping the same pattern on every square of a checkerboard, with each square representing a future chip.

Afterward, an etching machine engraves those patterns into the wafer, and more chemicals are deposited and baked onto the surface. This process repeats over and over as different patterns are laid on top of each other to create dozens of layers of transistors. Then the layers are connected to each other via copper wires to allow signals and power to travel throughout the chip.

The pods carrying the wafers rarely rest through this three-month process, traveling from machine to machine according to a preprogrammed route.

Belfi and the other engineers are there to ensure the machines don’t break down, a task that has grown more urgent as demand soars. Machinery in a typical chip factory, known as a fab, runs about 90 percent of the time, with 10 percent downtime for scheduled or urgent maintenance.

The most crucial — and expensive — pieces of equipment in the factory are the lithography machines that print the intricate designs on the wafers. The Malta plant has 20 of those; each costs roughly $100 million.

“When that machine breaks, it’s important you fix it pretty quickly,” says Peter Benyon, the factory’s general manager, who previously ran a GlobalFoundries plant in Singapore.

Last summer, as chip demand was soaring, one of the Malta factory’s lithography machines malfunctioned. Normally the equipment’s manufacturer, the Dutch company ASML, would dispatch an engineer to help fix it, but because of the coronavirus crisis, that wasn’t possible. So instead, a factory technician wearing an augmented reality headset connected with ASML engineers in the Netherlands so they could view the inside of the machine and oversee the repairs, Benyon said.

Getting machinery back online quickly can help a factory squeeze out more chips. So can cutting down on the number of faulty chips per wafer, Benyon said.

But dramatically boosting output means building new factories. By next year, chip makers will have started construction on 29 new fabs worldwide, according to SEMI, an industry association.

China and Taiwan will build more than half of these — eight each — followed by six in the United States, three in Europe and the Middle East and two each in Japan and Korea.

The United States manufactured more than a third of the world’s chips in the early 1990s, but production shifted to Asia as chip companies sought cheaper labor, and as Taiwan, South Korea and China began heavily subsidizing chip manufacturing.

Investors also pressured U.S. chip firms to focus on semiconductor design and to outsource manufacturing to Asia because of the huge cost required to maintain chip factories, according to Glenn O’Donnell, a tech analyst at market-research firm Forrester.

“It’s the byproduct of the short-term Wall Street mentality everyone has to buy into because you have to satisfy Wall Street,” he said.

Worried that the pendulum has swung too far, U.S. officials and lawmakers are now eager to rebuild domestic chip production with the help of federal subsidies. As chips become more central to weapons systems and to the economy overall, relying too much on Asia undermines national security, they argue.

“Having America have a robust supply of semiconductors is important to our national defense and our economic security,” then-Senate Minority Leader Charles E. Schumer (D-N.Y.) said during a visit to the New York factory last year, as he was promoting the federal funding bill.

GlobalFoundries is planning to expand the Malta factory’s output by at least a quarter if it receives some of the federal subsidy endorsed by the Senate, chief executive Tom Caulfield said in a February interview. Some of the company’s customers also are ready to invest to expand production to secure steady supplies, Caulfield said.

The goal is to double production capacity in Malta in the coming years with funding from the company, its customers and the federal government, spokesman Michael Mullaney said.

GlobalFoundries is using a similar combination of investment to boost output at its factories overseas; its planned $4 billion expansion of its chip plant in Singapore includes funding from Singapore’s government.

California-based Intel, meanwhile, has pledged to spend $20 billion to build two factories in Arizona. Intel CEO Pat Gelsinger has said the United States should aim to boost its share of global chip production back above 30 percent.

For now, U.S. manufacturers will have a hard time matching the low cost and sophistication of their Asian rivals, according to Steven Vogel, chair of the political economy program at the University of California at Berkeley, who has studied the chip industry.

“Taiwan and the Koreans are the best in the world when it comes to price and quality,” he said. The U.S. should take steps to rebuild its semiconductor manufacturing, but “there are huge efficiency benefits to the global supply chain,” he said.

The most advanced manufacturers — Taiwan’s TSMC and South Korea’s Samsung — both have announced plans to build new factories in the United States and are expected to be eligible for the federal subsidies should they become law.

Chip companies have emphasized the need for better STEM education and worker training to prepare the semiconductor manufacturing workforce.

“For us, one of the hardships is actually getting great folks because a lot of what you’re seeing here, there aren’t a lot of colleges that offer curriculum surrounding semiconductor manufacturing,” Belfi said.

There are several former auto mechanics on Belfi’s team, and some former Air Force mechanics who used to repair planes. “A lot of their outside passions are right up the block at Saratoga Speedway,” he said, referring to a nearby stock-car racetrack. “We look for technical skills, mechanical backgrounds.”

Other groups at the factory look for employees with degrees in materials science, chemistry or engineering, he said. GlobalFoundries offers apprenticeships that provide two years of on-the-job training to employees without a degree, and works with community colleges to help design chip-related curriculums.

Belfi himself studied mostly software and computer science in college but preferred building things, which he did for fun while coaching schoolkids competing on robotics teams.

“I was much more equipment-driven and liked to work with my hands,” he said.


Source : The Washington Post

Lagging but Motivated: The State of China’s Semiconductor Industry

Christopher A. Thomas wrote . . . . . . . . .

With advanced semiconductors key to powering a wide range of potentially transformative technologies, cutting edge computer chips have become a heated area of geopolitical competition for the 21st century. Despite their importance, semiconductors represent a rare area in which the Chinese economy is dependent on the rest of the world—rather than the other way around. Every year, China imports more than $300 billion of semiconductors, and most, though not all, major American semiconductor companies pull in at least 25% of their sales from the Chinese market.

This mutual dependence has benefitted the technology sectors in both countries. Every major Chinese technology company relies on U.S. chips: Tencent or Alibaba would not be the powerhouses they are today if they had relied on Chinese microprocessors during their formative years or had developed and manufactured their own. Many U.S. companies, meanwhile, have benefited from Chinese customers, markets, and innovations. The scale and cost reductions enabled by system and device manufacturing based in China and Asia more broadly has helped make information technology ubiquitous. Despite the harsh rhetoric on both sides of the Pacific, American semiconductor companies and their Chinese counterparts today are working together on hundreds, if not thousands, of product designs and joint technology development efforts.

Yet these collaborations have not prevented semiconductors from becoming a central faultline in tensions between the United States and China. In a post-COVID, post-Trump world, many in Washington would like to see the American economy less dependent on China and are exploring new restrictions on imports of Chinese hardware and exports of both cutting-edge semiconductors and the equipment required to manufacture them. Meanwhile in Beijing, Chinese officials are pursuing a clearly stated, though ambiguously defined, goal of “technology independence,” as articulated in the 14th five-year plan outlined last year.

But how to achieve that independence—and whether pursuing it makes sense in the first place—represents a question of profound uncertainty. As U.S. officials weigh their policy options in that regard, they first need to level-set on the state of the Chinese and global semiconductor industries, and how Beijing has approached its goal of building a domestic chip-making industry. Though it has made major advances, most segments of China’s semiconductor industry remain behind its foreign competitors, and its efforts to catch up face major economic obstacles. How the United States approaches its policy toward that industry will have major ramifications not only for the U.S. relationship with China, but also for the American semiconductor, systems, and internet services industries, which remain deeply intertwined with China.

Understanding the semiconductor industry

Although the United States is home to a majority of the world’s leading semiconductor firms, no country has true independence in the semiconductor value chain. The United States depends on critical foreign inputs and manufacturing capacity in the rest of the world. The manufacturing chain for any given semiconductor is extraordinarily complex and relies on as many as 300 different inputs, including raw wafers, commodity chemicals, specialty chemicals, and bulk gases; all are processed and analyzed by upwards of 50 different types of processing and testing tools. Those tools and materials are sourced from around the world, and are typically highly engineered. Further, most of the equipment used in semiconductor manufacturing, such as lithography and metrology machines, rely on complex supply chains that are also highly optimized, and incorporate hundreds of different companies delivering modules, lasers, mechatronics, control chips, optics, power supplies, and more. The “installed base” within a semiconductor factory today represents the cumulation of hundreds of thousands of person-years of R&D development. The manufacturing process that integrates them into a single manufacturing chain could represent hundreds of thousands more.

The types of products for which these manufacturing processes are designed are nearly as varied as the manufacturing inputs themselves. There are at least 20 major semiconductor product categories (from optical sensors to battery management modules to CPUs) and each category usually contains hundreds of different stock keeping units—distinct items for sale—for specialized applications. This complexity leads to a large market filled with myriad niches, in which specialized world-class companies have built defensible market positions through decades of targeted research and development.

Complexity also makes semiconductors a winner-take-all industry. The top one or two players in any given niche—whether a small one, such as furnaces, or a giant one, such as server CPUs—earn all the economic profits in that niche due to scale, learning efficiencies, and high switching costs for customers. It is rare to see newcomers break into these oligopoly positions. For instance, the market leader in graphics processing units (GPUs), Nvidia, invented the segment in 1999 and never relinquished its lead. While China has early-stage startups in the GPU segment, its market share is essentially zero. TSMC, based in Taiwan, was the first dedicated competitor in the foundry segment and has not relinquished its lead in its 33-year history. Indeed, SMIC, China’s leading competitor in the foundry segment, remains four or five years behind TSMC in technology, despite almost two decades of investment.

Beijing’s semiconductor strategy

Prior to the last half-decade, China spent more than 30 years and tens of billions of dollars to build a domestic semiconductor industry, showering its national champions with resources to compete with Western companies. Despite these investments, Chinese semiconductor companies make up a relatively small part of the global market. Chinese companies hold global share across the value chain at the following levels, according to data from IHS iSuppli:

Approximately 20% of fabless chip designers (the companies that design their chips and then contract out their physical production);
10% of global foundry plant capacity (the outsourced manufacturers that serve fabless chip designers);
Less than 1% of global integrated device manufacturing capacity (the companies that both design and manufacturer their own chips);
Less than 1% of electronic design software, semiconductor tools, and materials;
A less than 1% share in the most important end-product categories, such as the logic chips that are the brains of the internet or the advanced memory chips that major cloud vendors use to store trillions of photos and videos.
To counter its dependence on foreign suppliers of semiconductors, China announced a major new semiconductor policy in 2014. The “Made in China” policy, which launched the following year, included core technologies to semiconductors. The new semiconductor national policy contained two major innovations to previous industrial policy efforts: The first was to acquire technology from overseas via M&A; the second was to bring in “smart money” via private investors, such as private equity funds, to take the lead on investments. Over time, that policy has shifted toward a more traditional industrial policy model, with large manufacturing and R&D subsidies delivered to designated national champions. But with more than 50,000 Chinese entities registered as “semiconductor companies,” that investment is at risk of fragmentation.

Despite this effort to build up the Chinese semiconductor industry, the regional structure of the industry—based on the global distribution of market share using company headquarters location—was essentially unchanged in 2020 compared to 2014, and there has been no major shift to China in that time period. Chinese players remain decades behind in some of the most important manufacturing technology areas, such as lithography and the most advanced software design tools. The Chinese Semiconductor Industry Association estimates that to meet its policy goals, China will need to close a talent gap of roughly 300,000 engineers. In recent months, the Wuhan government stepped in to take over one cash-strapped manufacturer that had pledged to invest $20 billion in logic manufacturing, an example of the struggles facing the Chinese semiconductor industry.

Despite these setbacks, progress has been made. China’s share of back-end manufacturing, which is the labor-intensive process whereby processed semiconductor wafer is diced into individual chips, packaged, and outfitted with electrical connectors has nearly doubled since 2015 to 40% due to acquisitions. Market share of Chinese companies in fabless design has also nearly doubled, mainly via HiSilicon, the semiconductor arm of Huawei. These segments are more natural fits for China’s competitive advantages: Back-end manufacturing is a less technically challenging process that depends on low labor costs and operational efficiency; fabless design companies benefit from closeness to end customer applications (of which Chinese OEMs develop many) and have low barriers to entry due to widely available off-the-shelf design tools. There are now Chinese competitors in nearly every single stage of the value chain, including chemicals, materials, tools, and manufacturing, though some of their technology lags market leaders. China is turning its vast pool of venture capital and engineering talent to focus on this industry. In similar markets with lower technology barriers, such as MOCVD tools used for manufacturing LEDs, Chinese companies have developed competitive manufacturing equipment. While there are many potential outcomes of the Chinese semiconductor policy depending on tactics and the quality of execution by Chinese engineers, the Chinese semiconductor industry will undoubtedly become more competitive over the next ten years.

The 14th five-year plan appears to enshrine “technology independence” as a national strategy, though what independence means is not clearly defined. If we define technology independence as “self-sufficiency”—a fully in-country, PRC-controlled semiconductor supply chain that that serves all Chinese customer needs and does not infringe on the intellectual property rights of any global company—this is clearly not possible in the medium term. In the long term, it is also unlikely to be achievable. For one thing, the economics of an “only in China for China” supply chain do not work. Even if Chinese companies at each stage of the value chain win 80% of potential business from every potential Chinese customer, Chinese companies would collectively generate less than 15% of the industry’s overall R&D capacity—and likely less as prices in China tend to be lower, leaving less profit to re-invest in R&D. Such an indigenization strategy would still leave China behind the rest of the world: How can products developed with 15% of the world’s R&D compete with those from entrenched companies spending collectively far more? Of course, PRC government subsidies can and are closing that funding gap. But keeping such large-scale subsidies in place for the decades required to build the industry would likely generate a set of companies so dependent on government largesse that they may not be commercially viable.

As in the phone and PC industries, Chinese semiconductor companies would be better off focused on winning customers rather than winning government support, and by partnering with foreign firms to co-develop technology, build ecosystems, and pursue global customers. This is what several major Chinese consumer electronics companies, such as Vivo, Xiaomi, and Lenovo, as well as large telecommunications equipment manufacturers, have done. In many years, more than half of the sales of these leading device and system companies are outside China.

But as semiconductor technology has emerged as a point of competition with the United States, such partnerships have become more difficult. Since 2016, in response to concerns about Chinese tech firms’ involvement in human rights abuses and Chinese semiconductor companies’ links to Chinese military institutes, the U.S. government has added major Chinese consumers and producers of semiconductors to the entity list. These firms include semiconductor consumers such as DJI, ZTE, and Hikvision; targeted semiconductor producers include Huawei and SMIC. The companies on the list are generally ineligible to receive any item subject to Export Administration Regulations without a license provided by the Bureau of Industry and Security. The government has also tightened oversight of acquisitions of or investments in sensitive technologies such as semiconductors and has limited joint R&D and academic engagement between U.S. and Chinese companies, labs, and educational institutions. Taken together, these moves by the U.S. government have made the already difficult task of building a competitive Chinese semiconductor industry that much harder.

Despite the economic logic that would drive Chinese companies to globalize, many Chinese companies in the semiconductor value chain, from startups to established players, are quietly pursuing supply chain “indigenization” strategies. They are leveraging the current moment to focus on import substitution and are angling to remove American suppliers from their approved procurement lists or to become that alternative supplier. They encourage the Chinese press, government officials, and investors to view them as the “Chinese champion.” Executives in these Chinese firms are not making these decisions out of “patriotic spirit” or pursuing “national policy.” Rather, they are pursuing business-continuity goals, aiming to reduce risks from future export controls and entity listings (and, in many cases, hoping for support for a lucrative stock market listing). Since many of these executives have previously worked at large multinational corporations—and some even hold American passports—indigenization strategies are not ones that they would typically pursue. Yet the prospect of new and more extensive trade restrictions has led them to embrace more of an “in China for China” model.

The Chinese government is encouraging companies to follow this indigenization path. For example, the government has offered to provide insurance to Chinese companies to protect against faulty equipment or materials from Chinese suppliers and has also made manufacturing subsidies contingent on a commitment to use local suppliers. But make no mistake, the primary driver of the move toward indigenization are business decisions driven by businesspeople on the ground, rather than national policy.

How the Chinese and global semiconductor industry evolves from here depends largely on the strategic moves and the engineering execution of Chinese companies—but also on the policies of the incoming Biden administration. The move toward indigenization within China, coupled with widespread fears of American dependence on Chinese hardware and manufacturing, will pose a unique challenge for Biden’s foreign policy team. Along with their counterparts in Congress, the Biden administration will need to carefully weigh the benefits that come from U.S. semiconductor companies competing in one of the world’s largest markets against the risk of American-sourced technology being used to endanger national security; clarify the often blurry line between “primarily state-driven” and “primarily market-driven” operators in China; balance the benefits of having the best Chinese scholars studying in top engineering schools and collaborating with our national labs against the risks of industrial espionage and the infringement of intellectual property rights; and author a semiconductor industrial policy that continues to let the market decide winners and losers.

How the Biden team should approach those questions will the subject of an upcoming article. Yet even as the future of the Chinese semiconductor industry remains unclear, there is one undisputed outcome for now: The United States and its companies are losing influence and market share in the private Chinese technology sector. These vibrant, advanced, and innovative Chinese companies are increasingly determined to forge a new semiconductor ecosystem centered on China. The repercussions of this shift in mindset will reverberate far longer than any short-term benefits derived from using access to semiconductors as a negotiating lever in the larger context of U.S.-Chinese relations.


Source : The Brookings Institution



See large image . . . . . .

Source : TSMC

China Accelerates Production of 28nm and 14nm Semiconductors


See large image . . . . . .

The Covid-19 pandemic has exacerbated the pre-existing global shortage of semiconductor chips, which power everything from smartphones to vehicles. Geopolitical issues between the US and China have impeded international trade and driven the need for self-sufficiency, which China is pursuing.

China has been manufacturing integrated circuits domestically for many years, but it is now shifting its focus from fast production to high-quality end products. The partially state-owned firm SMIC (Semiconductor Manufacturing International Corporation) is already capable of producing both 14 nanometre (nm) and 28nm chips.

In 2019, China’s semiconductor manufacturing industry had an output value in excess of CNY750 billion, with this figure projected to be CNY884.8 billion in 2020. This swift progress has been made possible as China’s rate of domestic chip development has exceeded expectations, with SMIC achieving a breakthrough in 5G mmWave chips as well as its successful tape out of the N+1 process.

These technological improvements together with industrial policies have underscored China’s efforts to produce more chips domestically, with data from Analysys Mason showing that during 2020 Chinese semiconductor production rose by 16.2% – a sharp increase from 2019’s 7.2%.

This increase has been enabled by a number of factors, not least of which is the rapid development of China’s integrated circuit market. Over the past few years, this has driven progress in terms of both technology and industrial policy, closing the gap between China’s manufacturing processes and those of leading global semiconductor manufacturers. This will allow China’s development of integrated circuits to surpass other markets’.

A major factor here is that success in semiconductor development is self-perpetuating – the completion of new wafer fabs in mainland China has brought down the cost of domestic circuit production and thereby allowed it to expand, while the geographical advantages of local production have enabled development to flourish. The Chinese government has created a number of policies to support domestic production of integrated circuits – with a particular focus on creating higher quality products – acknowledging that the industry is a crucial part of national economic and social development.

At the heart of this are 28nm chips – seen as the bridge between the lower and higher ends of integrated circuit manufacturing capacity. Apart from CPU, GPU and AI chips that require relatively high power consumption, the majority of industrial-grade chipsets use 28nm or higher – they have become mainstream compared to 5nm or 7nm chips, and are used in a wide range of products including TVs, air-conditioners, automobiles, high-speed rail, industrial robots, elevators, medical equipment, smart bracelets and drones.

“The demand for China to move towards the mid-to-high end is pressing, one we have fully mastered the 28nm technology, we will be no longer stuck with most of the chip demand in the market,” according to Teng Ran, Deputy General Manager of CCIC Research Centre. “The production capacity of chip manufacturers in China on 28nm chips has been fully utilised, reaching at least 98%. This is the year with the highest capacity utilization rate in the past five years.”

Industry observers expect China to become self-sufficient in 28nm manufacture this year, paving the way for increased production of 14nm chips. SMIC was the first Chinese firm to produce this standard and mass produce the chips, joining the ranks of international firms such as Intel, TSMC, Samsung and UMC. The 14nm standard is used in around 65% of chips, and is expected to become the main technology used in mid-to-high-end semiconductors, with SMIC noting that the technology has significant potential in many fields – including high-end consumer electronics, high-speed computing, low-end AP and baseband, AI, and automotive applications. The 14nm chips that SMIC has produced can be used in the fields of 5G communications and high-performance computing. The firm has said that it expects to begin mass-producing the chips at scale from next year.

While China has made great strides towards self-sufficiency in chip manufacture, the industry must be globalised – the country’s independent research and development is aimed at better integration into the global market, rather than striking out separately and creating a new market. The chipset industry must be both capital-intensive and technology-intensive, but must also act as a collaborative system for global innovation.

“Our chip foundation is relatively weak. This is a fact that we must admit. However, against this backdrop, how to do it [increase production of high end chips] and meet the needs [of Chinese companies] can truly be a place that embodies the Chinese entrepreneurial spirit”, concludes Teng Ran.


Source : Developing Telecoms

The Future of China’s Semiconductor Industry

Paul Triolo wrote . . . . . . . . .

Over the past four years, the Trump administration—driven by growing concerns over China’s rise as a technological competitor and the coupling of its military and civilian industries—has ratcheted up controls on semiconductors and semiconductor manu­facturing equipment destined for Chinese end users. China hawks in the administration viewed American companies’ dominance of key semiconductor subsectors, particularly in areas such as electronic design automation and other tools needed to produce advanced semiconductors, as a key policy lever. They sought to use this lever to punish Chinese companies for specific types of activities deemed problematic by U.S. officials, and to push back on Beijing’s heavy subsidies to the semiconductor industry that some fear will distort market-driven global supply chains.

In response, Beijing, building on a series of policies in development for more than a decade, has over the past two years stepped up government support for the semiconductor industry across the board, via more traditional methods such as subsidies and through novel mechanisms designed to introduce market forces into the industry and attract domestic and foreign investment into this critical sector. This paper examines the challenges Beijing will face in trying to re­duce the dependence of Chinese technology companies on foreign sources of semiconductors and critical semiconductor manufacturing equipment, particularly from U.S. companies. For Beijing, this is a multiyear project, given China’s late entry into the industry, the huge and growing capital costs associated with manufacturing, shortages of trained personnel, and the fast-moving pace of innovation that keeps the cutting edge moving forward. Until recently, these factors have kept Chinese companies years and even decades behind leading global companies. To date, China’s progress has been highly uneven and inconsistent across the many subsectors that make up complex semi­conductor production value chains. This dynamic will continue to roil global supply chains in the semiconductor and broader technology sector over the coming decade, with significant collateral effects, both intended and unforeseen.

As this process unfolds, China will remain the largest global producer and export platform for electronic devices and systems, and thus also the world’s largest consumer of key inputs to advanced manufacturing in the semiconductor sector. China currently accounts for nearly one-third of global semiconductor demand, but domestic Chinese suppliers can provide for only around 10 percent of this demand. Chinese leaders have long been intent on reducing this gap, and much ink has been spilled designing plans and targets for domes­tic industry to begin winning market share across the many sectors that make up the semiconductor industry. Over the next decade, China will have to determine what level of domestic production of semiconductors and manufacturing equipment is appropriate and realistic, and at what cost that level can be reached.

Left to their own devices, Chinese companies competing in global and highly competitive sectors such as smart devices and telecommunications infrastructure prefer to use the best available technologies, including all types of semiconductors, regardless of country of origin. Even though Beijing is pushing for more domestic content, this dynamic will continue to hold for most Chinese firms with global market aspirations.

Senior Chinese leaders pushing the industry to be more innovative realize that China cannot realistically recreate, at home, the multiple and complex value chains that would be needed to achieve something like “self-reliance” in semiconductors. A leading Chinese semiconductor industry advisor, Professor Wei Shaojun, made this point in a speech at an industry conference in late 2020, cautioning that “the idea of ‘All Made in China’ is very hot, and is getting too hot.”1 Nevertheless, political leaders and industrial planners will continue to pull out many stops to achieve significantly lower levels of dependence.

Beijing Doubles Down on Core Technologies

Chinese industrial planners have long focused on the development of domestic capabilities in so-called core technologies, with semiconductors and software topping the list. Chinese companies were late in getting started on the design and manufacturing of semiconductors. During the period from roughly 1990 to 2000, the focus was on state planning, with the national and local governments backing large semi­conductor manufacturing efforts. Beginning with the 1991 Eighth Five-Year Plan (FYP) document, semiconductors have continuously been a focus of state planning—in the Eighth FYP, development of the industry was termed a “main task” of the state. These efforts ultimately did not produce major breakthroughs in domestic capabili­ties; they did, how­ever, begin laying the groundwork for eventual progress in some areas.

From roughly 2000 to 2013, state industrial planning continued, now coupled with a gradual internationalization of the Chinese semi­conductor sector. This period saw the establishment of China-based foundries—by far the most important being Semiconductor Manufacturing International Corporation (SMIC)—that were managed largely by foreign semiconductor industry veterans, many from leading Tai­wanese firms. It also saw the opening to foreign direct investment in the sector, a focus on developing Chinese companies that would come to play leading roles in areas such as assembly and testing, and the rise of China’s fabless design firms.

Critically, during the latter part of this period, Chinese semiconductor design houses began leveraging the advanced manufacturing capabilities provided by Taiwan foundry leaders Taiwan Semiconductor Manufacturing Corporation (TSMC) and United Microelectronics Corporation (UMC). The role of Taiwan-based companies in China’s rise as a semiconductor power would only grow. Chinese firms also began working with other foundry leaders such as GlobalFoundries and South Korean giant Samsung. This was because SMIC, despite making steady progress up the technology chain, found itself—and continues to find itself—several generations behind the leading edge of semiconductor manufacturing.

With the ascension of Xi Jinping to Party General Secretary in 2014, and Xi’s emphasis on turning China into a cyber superpower, including a renewed emphasis on developing core technologies, China’s approach to the sector changed. With very focused industrial policy choices becoming the norm, China introduced new and inno­vative approaches to developing domestic capabilities across the semi­conductor supply chain. The State Council’s release of its National Integrated Circuit Industry Development Guidelines in 2014,2 and the establishment of both the National IC Investment Fund in 20143 and the release of the Made in China 2025 (MIC2025)4 initiative in 2015, marked a new turning point in Beijing’s approach to pushing the sector toward reduced dependence on foreign suppliers.

The implementation of the strategy under the Guidelines fell to the National IC Industry Leading Small Group, led at the time by Vice Premier Ma Kai. Critically, the group is advised by a consulting com­mission (referred to as the A-Team), which includes twelve experts from government, academia, industry, and the government-backed national investment fund bodies. These experts, such as Wei Shaojun, Wang Xi (president of Shanghai Institute of Microsystem and Infor­mation Technology), Zhao Weiguo (chairman of Tsinghua Unigroup), and Xu Xiaotian (executive vice president of the China Semiconductor Industry Association or CSIA), are deeply involved with global semiconductor industry groups. Wei, for example, is a Chinese delegate to the World Semiconductor Council and a frequent participant in Global Semiconductor Alliance conferences. He is viewed as an expert particularly on the details of the market-driven global semiconductor industry value chain. It is not surprising that Wei is critical of a Chinese go-it-alone strategy in the semiconductor sector.5

In October 2015, a companion document to the MIC2025 program was released, drafted by engineers at the Chinese Academy of Engi­neering (CAE).6 It included details on the desired domestic market share percentage for key semiconductor subsectors. The document received significant attention from Western analysts, who frequently refer to it as an example of Beijing’s industrial policy goals.7 The document was written without reference to the market forces driving the industry, however, and its 2025 targets for domestic production are aspirational. Chinese companies are nowhere near being capable of meeting them. In addition, the M&A efforts fueled by the National IC Investment Fund during this period led to a major pushback against Chinese investments in the sector late in the Obama administration. Obama’s commerce secretary, Penny Pritzker, called the Fund an effort by China to expropriate the global semiconductor supply chain.8 Perceptions of the intent and aspirations of both the Fund and MIC2025 also contributed to the substantial pushback against Chinese semiconductor and technology companies under the Trump administration. Chinese industrial policies to promote domes­tic capa­bilities, including the Fund, feature prominently in the Section 301 report by the U.S. Trade Representative (USTR) on China’s trade policies in March 2018, which became the center of the U.S.-China trade dispute.9

The latest phase of China’s semiconductor industry arguably began around 2018, driven by new pressures from the outside—par­ticularly the U.S. reaction to the National IC Investment Fund and MIC2025, coupled with the weaponization of export control policies by the Trump administration. These restrictive policies cut several ways: they were first launched against leading telecommunications vendors ZTE and Huawei, and later against Chinese companies in other sectors such as artificial intelligence and supercomputing.10 They had the effect of re-galvanizing Beijing’s efforts to inject funding and innovative capacity into China’s growing semiconductor sector. In addition, U.S. efforts to restrict SMIC from acquiring advanced manufacturing equipment become clear during 2018–19, and industrial planners and senior Party officials became increasingly determined that China would have to adopt a two-track long-term policy to reduce China’s dependence on semiconductors and semi­conductor manufacturing equipment (SME) to avoid being subject to what they viewed as U.S. attempts to politicize complex global supply chains for semiconductors, critical design tools, and SME.

The State of China’s Semiconductor Sector in 2021

Before looking at the new two-track approach, it is necessary to survey and assess the current state of China’s semiconductor indus­try. Any attempt to determine where China’s semiconductor industry stands in relation to global leaders requires breaking down the sector along its separate lines, usually either by product line or, for SME, by different critical supply chain segments. In some areas, Beijing has long been concerned about overdependence on foreign sources for more than 80–90 percent of supplies, with little change in the num­bers over the past decade.

Two significant questions for Chinese industrial planners will continue to be salient: (1) what is the level of market share that Chinese companies can realistically achieve in key market subsectors, each of which has its own unique barriers to entry and technology road map; and (2) what is the most effective role of government in fostering companies that can successfully develop and compete in the many sectors where China is heavily dependent? Technology road maps can also change rapidly due to market conditions, and Chinese planners have been unable to anticipate major changes in most mar­kets. The most successful Chinese companies have been those in highly competitive, market-driven segments, such as Huawei in smartphones and telecommunications equipment. High levels of innovation are required to compete at the cutting edge of both handsets and infrastructure.

Before looking at specific pieces of the China semiconductor sector puzzle, it should be noted that China has one unique conglomerate, Tsinghua Unigroup, which includes subsidiaries that occupy places in a number of key subsectors. Unigroup’s five primary sub­sidiaries include memory producer Yangtze Memory Technologies (YMTC), fabless smart device designer Unisoc, smart card and mem-ory company Guoxin Microelectronics, and server, storage, and related technology provider H3C Technologies. Tsinghua Unigroup chairman Zhao Weiguo has attempted to drive the company’s devel­opment of core semiconductor design and manufacturing capabilities via mergers, acquisitions, and patent licensing.11

The company is viewed as a major piece of MIC2025 and Zhao is seen as a mover and shaker in the semiconductor sector. Industry observers believe Tsinghua Unigroup has close ties to senior Chinese leaders, including General Secretary Xi Jinping, and at least until last year, has benefited from government policies and financial subsidies from the National IC Investment Fund.12 There has been considerable discussion lately about Unigroup and Zhao’s star falling, given the struggles the company has had with mergers and acquisitions, master­ing a variety of technologies, and recent financial issues, including defaulting on several major bond issuances.13 Tsinghua Unigroup will continue to play an important role in China’s semiconductor ambi­tions, but it is becoming increasingly unclear whether the firm’s busi­ness model will facilitate breakthroughs in capabilities across key subsectors.

First, in terms of foundry capabilities—that is, manufacturing the designs of fabless companies—China has only two companies that are globally competitive: SMIC and Hua Hong. In 2019, these firms held just 5 and 2 percent, respectively, of global foundry market share, with TSMC (54 percent), Samsung (20 percent), Global Foundries (9 percent), and UMC (8 percent) holding higher shares.14 There are no Chinese integrated device manufacturers (IDMs), as most companies have or are moving to the foundry model. Nevertheless, under the influence of the Outline and the National IC Fund, the Chinese manufacturing sector has seen an average compound growth rate of nearly 25 percent since 2014.

In the related semiconductor design segment, China has made huge strides over the past fifteen years. In 2019, China’s design sector reached a level that surpassed Taiwan, making China the second-largest design industry cluster after the United States globally. Chi­na’s share of semiconductor design went from 3.6 percent in 2004 to nearly 43 percent in 2019.

On the product side, broadly speaking, Chinese companies have only really been able to gain significant market share in the mobile processor space. For most types of logic semiconductors, including central processing units (CPUs), increasingly important graphics processing units (GPUs), and other specialized semiconductors such as field programmable gate arrays (FPGAs), domestic Chinese com­panies hold very small market shares, in some cases less than 1 percent. After many years of government subsidies and other pref­erential policies, domestic firms such as Loogson and Tianjin Phyt­ium have a very small but growing market share, primarily in the domestic and not the global market, the latter being dominated by U.S. firms Intel and AMD.15 Phytium, in particular, part of the so-called PK system (where the firm’s CPUs are paired with domestically developed operating systems), could become a key player in efforts to replace Western CPU designs. Recently there have been a number of new entrants to the field, including e-commerce giant Ali­baba and start-up Zhaoxin, which are competing in the CPU and embedded CPU sectors. Chinese companies have so far not made a significant dent in the market share for GPUs, which are dom­inated by U.S. giants nvidia and AMD.

For mobile application and communications processors, Chinese firms including Huawei, Xiaomi, and Unisoc have made significant inroads into global market share, driven by meaningful advantages in design capabilities, coupled with the ability to manufacture at TSMC. In particular, Huawei’s design arm HiSilicon, prior to U.S. actions in 2020 that cut the firm off from manufacturing at TSMC, had moved into a position where some of its mobile handset and 5G infrastructure semiconductor designs were on par with Western leaders such as Qualcomm. HiSilicon was also pushing its semiconductor design capabilities across Huawei’s product lines, including for server and AI chips. Tsinghua Unigroup claims it is the world’s third-largest design­er of commercial mobile phone semiconductors and holds around 20 percent of the global SIM card market.

In the memory sector, Chinese firms are just starting to make significant inroads into the global market, after massive investment via the National IC Fund and local versions of the fund, and the failure of a number of ventures. Domestic firm GigaDevice, for exam­ple, accounted for nearly 20 percent of global market share for NOR flash memory in the first half of 2020 according to industry sources. Major Chinese memory players YMTC, a maker of 3-D NAND, and Changxin Memory Technologies (CXMT)—formerly called Innotron—a DRAM company, are both poised to break into global mar­kets in 2021 using fairly advanced manufacturing processes. YMTC is using an advanced 128-layer process, and expects to reach high-volume commercial production in mid to late 2021.

In a major new market segment, dedicated or optimized AI processors, there are a number of Chinese start-ups—including Cambricon, Horizon Robotics, Bitmain, Thinkforce, Enflame, and Novumind, and bigger players including Huawei, Alibaba, Tencent, and Baidu—that could become major players over the next five years. These are primarily edge AI processors, designed for specific tasks such as image recognition or natural language processing, and applica­tions in automotive, surveillance, or smart cities.16 A number of start-ups are also pursuing AI-optimized semiconductor development for cloud-based applications, attempting to challenge U.S. companies such as nvidia and start-ups such as Cerebras. Larger players such as Alibaba and Baidu have some advantages in this space, but will have a hard time competing both on the hardware performance side and in the software development ecosystem where U.S. companies generally have robust offerings.17

On the design tools and manufacturing equipment side, Chinese firms have even less market share in most areas. This sector breaks down into (1) electronic design automation tools (EDA) required for the design of cutting-edge chips and (2) manufacturing equipment, including a wide range of tools from lithography, etching, and mask writers to critical process control, metrology, and inspection tools. In addition, Chinese firms lag in the development of critical materials used for semiconductor production, such as IC-grade polysilicon, materials for advanced extreme ultraviolet lithography (EUV), and photoresists. The global EDA tools sector is dominated by U.S. firms Cadence and Synopsis as well as Mentor, a Siemens subsidiary. Chi­nese domestic leader Empyrean is focused on analog design tools, and claims its tools are used by leading players such as HiSilicon and UniSoc. Empyrean is also pushing into the digital/logic IC design tool arena, and its tools here have also been used by HiSilicon and Unisoc. In general, however, Chinese EDA tool developers are far behind the global leaders in the types of processes and IP they can support.18

It is in the area of advanced lithography tools that China is likely to face the most challenges. Chinese leader Shanghai Micro Electronics Equipment Co. (SMEE) is producing equipment only capable of handling critical and noncritical layers at the ninety-nanometer node, well behind industry leaders such as ASML, which are already sup­porting volume production at the seven- and five-nanometer nodes. Moreover, the U.S. government has pressured the Dutch government into blocking ASML from shipping EUV equipment to SMIC that had been contracted for in 2017.19 Details of the transaction and U.S. government policy on EUV technology remain unclear: there is U.S. intellectual property that is part of ASML’s EUV systems, and under the Trump administration, efforts to control advanced semiconductor manufacturing equipment have moved forward, but still have not been clarified.

Several other Chinese firms are beginning to move up the technology value-added chain in the SME sector, such as Advanced Micro-Fabrication Equipment (AMEC) and Naura Technology Group, which manufacture a range of semiconductor production systems, including oxidation, epitaxy, laser annealing, etching, and chemical vapor deposition (CVD) systems. But these firms hold only small market shares: in etching, Naura holds 1 percent as of 2019 and AMEC less than 1 percent, and Naura holds 1 percent in epitaxy. In most of these sectors, U.S. companies hold 70 or 80 percent market share.

Finally, a key area where Chinese firms have been able to grab significant market share is in the outsourced assembly and test (OSAT) sector. Here, companies such as JCET, Tongfu Microelectronics, and Tianshui Huatian Technology hold a significant share of the global OSAT market, which is largely dominated by Taiwan-based firms.

Beijing’s Semiconductor Strategy: A Tale of Two Tracks

To address these challenges and shortfalls, Beijing is constantly working to determine the most appropriate role for the state, at both the national and local levels, in helping to create a policy atmosphere that optimizes the potential for developing successful companies that can compete with leading players globally. In addition to commercial gains, Beijing’s goal is to reduce China’s dependence on foreign sources of semiconductors and SME and the political risks of such dependence going forward.

As Beijing surveyed the semiconductor industry landscape in the first half of 2018, along with the geopolitical situation, several things became clearer. The U.S. government was intent on using a range of policy instruments to punish Chinese firms for various reasons, including supply chain issues, alleged human rights violations, Chi­nese firms’ links to the Chinese military, and Beijing’s military-civilian fusion initiative, which Xi had begun touting again as a critical part of his efforts to modernize the People’s Liberation Army (PLA). At the same time, Beijing’s efforts to push domestic companies to buy Chinese-made semiconductor and other high-tech products such as software had shown mixed results, and in some critical sectors, such as SME, Chinese firms were far behind industry leaders.

Track 1: sharpen and provide resources for internal policies to push adoption of domestically produced technologies, including semiconductors. After a critical meeting in April 2018, Beijing launched a new project, centered on the concept of developing “secure and controllable” (安全可靠) supply chains. Details of this initiative are difficult to find in public media, and a number of references to the program have been pulled from various Chinese financial and industry-related web­sites. The project, designed to operate on a five-year time horizon (to 2023), under guidance from a senior Party organization, has the goal of replacing foreign—mostly U.S.—components and systems with domestic hardware and software in critical information systems. The goal of the project appears to be much more ambitious than the “De-IOE” (IBM, Oracle, and EMC) campaign that was launched after the 2012 disclosures by Edward Snowden about U.S. company involvement with U.S. intelligence agencies.

While the focus of the secure and controllable effort was initially on system vendors, as a result of U.S. export control pressure on the supply chains of ZTE and Huawei, officials in Beijing are now mandating that companies, particularly large state-owned telecommunications companies, begin to phase out reliance on U.S. semiconductors and other critical components. The effort does not apply to semiconductors from Japanese, South Korean, European, or Taiwan-based companies. The project also likely includes efforts to push the use of domestic alternatives to the x86 CPU architecture running a Windows operating system. A major Chinese bank in early 2019 indicated it was using a CPU designed by Tianjin Phytium (Feiteng) and the indigenously developed Kylin operating system for all its domestic and overseas offices.20 Other Chinese companies such as Alibaba, Tencent, Lenovo, and Inspur also announced server prod­ucts based on the “PK system.” The China Electronic Corporation (CEC), which owns Tianjin Phytium, continues to push the PK system as a completely indigenous, secure, and controllable alternative to foreign hardware and software. In December 2020, industry media reporting claimed that over four hundred “core” enterprises were working with the PK system to establish an indigenous IT industry and to promote the ARM architecture for corporate net­works, IoT, cloud computing, and big data applications.21 The latest Phytium chips are all ARM-based, and in July 2020 the company announced the Tengyun 2500 series, designed for servers. The Phytium roadmap for CPUs extends to 2022 when the firm claims it will be manufacturing at the five-nanometer node—and presumably this means the manufacturing will be done at TSMC.22

The long-term viability of the secure and controllable strategy remains unclear, but the elements of an alternative domestic stack that includes the PK system appear to be gaining traction, with internal government mandates to swap out Western semiconductors where possible, accelerating industry acceptance. But it remains unclear whether the PK system is capable of handling mission-critical work­loads for key industries such as the financial sector.

Track 2: develop new approaches to financing for the sector and new economic policies to incentivize investment in domestic capabilities. The second track that began to take shape in 2018 included initiatives undertaken by senior leaders and industry planners to introduce market mechanisms into the broader high-tech sector, and specifically the semiconductor industry. At the same time, there was a realization that new policies would be needed to further incentivize companies to step up R&D efforts and to provide support for the industry going forward in order to help bridge what is likely to be a long period of catch-up innovation.

First, in mid-2019, at the behest of Xi Jinping, financial market authorities rammed through approval for a new high-tech stock board in Shanghai, the so-called STAR market. The long-awaited board provides a new mechanism for funding promising tech start-ups, including in the semiconductor industry. It has also been used by established companies, such as SMIC, to raise money. In July 2020, SMIC raised nearly $8 billion before its IPO on the STAR board.23 In addition, as of July 2020, more than forty semiconductor companies were already listed or were applying for listings on the STAR board, suggesting that regulatory authorities were prioritizing an industry that Xi had emphasized as a key part of reducing China’s foreign dependence on core technologies. The long-term success of the new Shanghai STAR market, however, hinges upon capital and technology market dynamics through which innovative tech companies, mature institutional investors, and sophisticated regulators interact with each other and are able to establish a sustainable and healthy ecosystem.

In August 2020, Beijing unveiled a new set of measures to promote the domestic semiconductor industry, including significant tax relief for companies engaged in semiconductor manufacturing, design, or design-related software.24 The new State Council policy specifies that qualifying semiconductor projects and enterprises using processes at the twenty-eight-nanometer or more advanced nodes, and which have been in operation for more than fifteen years, are exempt from corpo­rate income taxes for up to ten years. Meanwhile, projects at the sixty‑five-nanometer to twenty-eight-nanometer nodes have a five-year tax-free period and will enjoy a 50 percent discount on the corporate rate for the following five-year period. For companies in the semiconductor design, packaging and testing, and other manufacturing equipment sectors, along with semiconductor materials and software (EDA) firms, their first two years of profits will be exempt from corporate income taxes, while these companies also get a significant discount on tax rates for the following three years. Currently, only SMIC and Hua Hong can produce products at the twenty-eight-nanometer level. While some industry observers did not think these measures would have a major impact, they were an improvement and reflect the importance Beijing is giving to the industry.

On October 29, the Party’s Central Committee released a communiqué summarizing its deliberations during the Fifth Plenum.25 Technology and innovation were at the center of the discussions, in terms of the impact that they would have on balances of power and economic growth, as well as on the efforts that China would need to undertake to remain at the forefront of global com­petitiveness. The Plenum document emphasized that achieving major breakthroughs in core technologies—meaning semiconductors and software—was a top priority for the country’s long-term development roadmap, aiming for completion by 2035. It also highlighted innovation as the driving force for modernization and technological self-reliance as the cornerstone of development in the dual circulation era. Xi, in speeches around the Fifth Plenum and discussion of the Fourteenth Five-Year Plan, and at economic and trade-related confer­ences, stressed the need for China to become self-reliant in core technologies.

The communiqué coined the new term “technology self-reli­ance” (科技自立自强)26 to describe Xi’s vision for a domestic-tech­nology-driven economic growth model amid decoupling pressure and weakening global demand.27 It also indicated that Beijing is likely to keep semiconductor technology development as its top priority for the foreseeable future. Next-generation technologies and digital infra­structure are of primary importance for the nation’s overall strategy. Noticeably, the statement did not mention controversial policies such as MIC2025 or the military-civil fusion policy, part of continuing efforts in China’s state-run media to downplay domestic industrial policies. The draft plan typically goes through two more rounds of internal comments before final publication, scheduled to take place in the spring of 2021.

The pathway to self-reliance in sectors such as semiconductor manufacturing equipment will be long and difficult for China. In a key illustration of the challenges China will face, as of late 2020 Huawei, under pressure from U.S. government export control ac­tions, was devising plans to equip a semiconductor manufacturing plant in Shanghai with non-U.S. tools, aiming to create a new supply chain free from the threat of U.S. export controls.

The firm’s semiconductor manufacturing partner, Shanghai IC R&D, is backed by the Shanghai municipal government and partially owned by state-run Huahong Group, which also has shares in domestic chipmakers Huahong Grace and Shanghai Huali Microelectronics Corporation (HLMC), two other leading domestic foundries.28 But IC R&D will start with the production of low-end chips at forty-five nanometers, an older generation of technology used by global industry leaders fifteen years ago. The company aims to meet Huawei’s goal of making twenty-eight-nanometer chips by the end of next year, to allow the telecom equipment major to incorporate home‑grown semiconductors into smart TVs and other Internet of Things devices. Assuming the facility can achieve this step, Huawei would like to move to more advanced twenty-nanometer semiconductors by the end of 2022. These chips could be used in the firm’s 5G infrastructure products, but it remains unclear how Huawei can maintain performance metrics for 5G systems using the less capable hardware.

Even if IC R&D is able to deliver products according to this aggressive timetable, domestic semiconductor manufacturing capabili­ties will not be sufficient to support Huawei’s smartphone business lines, which require the use of semiconductors at cutting-edge tech­nology nodes. The replacement of U.S.-origin semiconductors and cutting-edge devices manufactured in Taiwan using U.S. technology will be technically and financial challenging. This will be the case even with government policy and financial support, and other factors such as the fact that Chinese firms have higher tolerance for production cost and operational inefficiencies.

Finding adequate replacements for U.S.-origin manufacturing tools will be even more difficult. The Shanghai project will begin production with second-tier equipment from Chinese suppliers such as AMEC and Naura, together with secondhand, high-end tools made by foreign manufacturers available on the market. Industry experts believe that it will take years to substitute for the suite of cutting-edge foreign semiconductor manufacturing equipment required to produce devices at more advanced processing nodes.

Technological capability will be a major limiting factor going forward, despite all the other strategies Beijing is deploying to provide the domestic semiconductor industry with virtually unlimited financing and a very favorable policy environment. The dominant position of U.S. and Japanese companies in the manufacturing sector, for example, is the result of very strong and long-standing customer relationships. This means equipment suppliers codevelop manufacturing processes across multiple generations of increasingly advanced products, creating a virtuous cycle of knowledge development and technology iteration that results in considerable accumulations of engi­neering and systems-based knowledge that is very difficult to repli­cate or obtain through illicit means. While Chinese companies can gain in some limited market segments, it would be difficult for domestic manufacturing firms to rely on Chinese-developed equipment for scaled and sustainable commercial operations.

Each of these measures will take time to begin having an effect on domestic capabilities across all segments of the semiconductor indus­try. This is the problem facing Chinese planners: fostering a sizable group of capable companies that can master complex technologies across multiple sectors in a highly competitive global environment, where no single country has been able to go it alone.

The Industry’s Future in China, the United States, and Taiwan

As Chinese industrial planners and industry executives look forward to 2021 and beyond, they will have to grapple with a domestic industry that faces profoundly different challenges than it did in early 2018. These issues highlight how closely the American, Chinese, and Taiwanese semiconductor sectors are intertwined, via supply chains, manufacturing relationships, and close financial and personnel link­ages. Primary among Beijing’s challenges will be dealing with what appears to be the freezing of domestic manufacturing capabilities at around fourteen nanometers, and potentially further restrictions on Chinese companies’ ability to design advanced semiconductors and manufacture them in Taiwan. In addition, Beijing will be looking to determine if alternative approaches to semiconductor design and manufacturing, such as the open-source risc-v architecture,29 will mature in time to offer Chinese companies another path to reducing dependence on proprietary approaches subject to U.S. export control restrictions.

The new Biden administration could also review U.S. policy on allowing the sale of advanced manufacturing equipment to Chinese end users such as SMIC. Late in its tenure, the Trump administration pressured the Dutch government to block the transfer of advanced EUV lithography equipment to SMIC. After much debate and indus­try pushback, China hawks succeeded in adding the firm to the Entity List restricting exports in late 202030 and included the firm on the Department of Defense Section 1237 List.31 Additionally, a November 2020 executive order bars U.S. persons from investing in companies on the list.32 As of early 2021, SMIC and other Chinese companies were being removed from financial indices and U.S. exchanges an­nounced plans to delist the named companies.

Administration critics and some in the Biden team criticized the interagency process that led to the development of the Section 1237 List and might consider rolling back the executive order. But the political climate in Washington—and the linkage of restrictions on SMIC to the company’s alleged support for China’s military—will make it politically difficult. On the other hand, SMIC has denied providing support for the military, and elements of the U.S. semiconductor industry are likely to push for revisiting all aspects of the U.S. treatment of SMIC. The outcome of this debate will have a huge impact on the future development of China’s semiconductor manufacturing equipment. While the wording of the November 2020 Entity List action implied the presumption of denial for equipment unique to manufacturing at nodes below ten nanometers, SMIC is currently not manufacturing at that level, suggesting limited impact on current production. But the Entity List action could, depending on how licensing is implemented, have a major impact on the firm’s ability to do research on advanced manufacturing techniques needed to continue moving up the technology value-added chain.

If the Biden administration continues to block shipment of EUV equipment to SMIC, China’s domestic manufacturing industry will be essentially frozen at around fourteen nanometers. SMIC could push the capabilities of its existing dense ultraviolet (DUV) lithography equipment down to seven nanometers in some layers, but this would come with major challenges in maintaining commercial yields.33 In addition, the swirling regulatory uncertainty around SMIC as of late 2020 had also led to internal management problems, all of which threaten to undermine the ability of the firm to attract critical talent going forward. If SMIC is forced to refocus on less advanced tech­nologies, which seems probable, it will likely be very hard for the firm to attract the management and engineering personnel needed to remain a global player in the industry.34 Finding enough talented managers and engineers has been a general problem for China, but the shortage of managers with deep industry knowledge will be particularly difficult to overcome if U.S. pressure on the industry remains high.

In addition, the Biden administration could decide to add more semiconductor manufacturing technologies to the list of so-called foundational technologies called for under the 2018 Export Control and Reform Act (ECRA). While the Trump administration has been unable to agree on adding technologies to the list, the Biden administration might prefer to go this route, rather than targeting individual Chinese companies, to impose broad restrictions under Commerce Department authorities. It will probably take some time for the Biden team to develop its overall China strategy before considering a new approach to export controls for companies like SMIC and Huawei.

Finally, the Biden team will have to grapple with the argument raised since 2016 by Pritzker and others,35 including in the U.S. semiconductor industry, that the United States should restrict exports of SME and design tools to Chinese semiconductor firms that are heavily subsidized by Beijing because, over the long term, heavy subsidies will lead to market distortions that undermine trusted U.S. suppliers of semiconductors.36 This argument was made in the 2018 Entity List action against Fujian Jinhua, a DRAM manufacturer that was also indicted for participation in stealing IP from U.S. memory leader Micron.37 The same argument could be used to add virtually all of the current fabs under construction or operating in China. A U.S. government contractor report on YMTC released in early 2021 made similar arguments for action against the company, including placement on the Entity List and/or on the Department of Defense Section 1237 List.38 This issue may be one of the most complex to confront the Biden administration in determining what policies to implement to ensure both U.S. semiconductor industry innovation and long-term competitiveness, as well to blunt the impact of massive state-led industrial development of the semiconductor industry in China.

Foreign Pressure Will Not Deter Beijing

Meanwhile, Beijing’s domestic policy push under the secure and con­trollable program will continue to mean that local suppliers are favored and a localization trend will continue. For example, local suppliers are now able to supply some products in all manufacturing segments, except for advanced lithography tools. The lower cost of domestic tools and the improving levels of technology and quality will mean that local manufacturers such as SMIC, YMTC, and Hua Hong increasingly look to domestic suppliers.

While this process was underway, the geopolitics of semiconductors changed again for Beijing in mid-2020, when China hawks in the United States stepped up efforts to further restrict the semiconductor supply chain of Huawei. For the first time, Taiwan was dragged headlong into the fray, marking another major turning point. The United States in May 2020 took steps to cutoff Huawei from its critical manufacturing base in Taiwan, particularly from TSMC. The modification to the Commerce Department’s foreign direct product rule (FDPR) extended U.S. export controls globally to any company manufacturing semiconductors on behalf of Huawei.39 This led to TSMC’s decision to cut off its relationship with Huawei as of September 2020. China’s industrial planners likely fear that the FDPR change could be applied to other Chinese companies. The ramifications of these actions are still unclear, but they have the potential to generate major geopolitical conflict between the United States and China under the Biden administration.

The inclusion of Taiwan in the U.S.-China technology conflict under the Trump administration has so far not resulted in a major response from Chinese authorities. Any further pressure on Taiwan and Taiwanese companies, however—including perceptions that the U.S. government was forcing Taiwan to choose sides in a red-versus-blue supply chain contest—could provoke further reaction from Beijing and heighten the potential for a geopolitical confrontation centered on semiconductors.40

Much will depend on how the Biden team decides to approach these complex issues. Some U.S. semiconductor industry officials have argued that the types of policies pursued by the Trump administration have forced Chinese companies to consider excluding U.S. technology from their supply chains over the long term, and that this will eventually erode U.S. market share and the profits that U.S. companies could plow back into R&D. Industry observers also argue that the U.S. government needs to be more sensitive to the impact of policies targeting Chinese firms that are already threatening the sustained integration of Chinese technology firms into essentially U.S.-company-led semiconductor ecosystems.41 Biden administration officials will likely be more receptive to this argument than the China hawks on the Trump national security team were.

Still, balancing economic and national security policies in the semiconductor arena will be a challenge. Rolling back select Trump-era restrictions on Chinese firms would benefit—and be supported by—some parts of the industry, but such measures could run into opposition in Congress and other quarters. The degree to which the Biden team chooses to weaponize U.S. company advantages will be a key issue in determining where China’s semiconductor industry, across almost all segments, is headed.

For Beijing, the only choice is to continue building a favorable regulatory environment for the domestic semiconductor industry, while trying to avoid overheating and wasteful investment, and striv­ing to enable a long-term, sustainable semiconductor sector. Beijing has so far not shown any inclination to reduce the heavy subsidies to the sector, though as of late 2020 Chinese planners were concerned about overinvestment and were taking a hard look at the financial viability of major projects.42 The National Development and Reform Commission (NDRC) expressed concern in late October 2020 that the semiconductor sector had experienced “chaotic” development and many companies had “blindly entered into projects.” The NDRC indicated it would be working with financial institutions and banks to coordinate assessments of the viability of semiconductor projects going forward.43 Some industry observers have suggested this could be the start of a major rethink in Beijing of how to finance development of the sector.44

Several major semiconductor projects as of late 2020 were under financial pressure and had either folded or were on the verge of being restructured. The collapse of Wuhan Hongxin Semiconductor Manu­facturing (HSMC)45 in late 2020 may be a harbinger of trouble ahead for the industry, which already has seen massive overinvestment chasing too few viable projects. In addition, semiconductor major Tsinghua Unigroup’s financial difficulties, already impacting several high-profile memory projects, are likely to continue, and could require a major restructuring of the conglomerate’s semiconductor businesses.46

Another critical issue Beijing will be grappling with in 2021 is how to handle mergers and acquisitions in the semiconductor arena that could impact China. The proposed acquisition of ARM by U.S. giant nvidia, for example, could be rejected by China’s antitrust regulator. Many Chinese semiconductor design companies depend on ARM cores and architecture, and U.S. restrictions related to Huawei affect­ed ARM products. Chinese domestic companies are probably lobby­ing Beijing to reject the deal, concerned about the United States gaining dominance of another key piece of the supply chain that could potentially be weaponized.47

The tension between Beijing’s preference for managing the domes­tic industry and the market-driven nature of the global industry will continue to produce inefficiencies and wasteful investments. In the 2019–20 period, a number of major domestic manufacturing projects collapsed or had to be restructured, and there will likely be a major shake-up in the industry in 2021, with planners in Beijing hoping that some strong companies will emerge out of this process.

Regardless, for the foreseeable future, the semiconductor industry will remain a critical sector, a high stakes geopolitical issue, and a source of tension between Beijing, Washington, and Taipei. Many have argued that U.S. pressure on China in the semiconductor arena will ultimately mean that domestic companies step up and overcome the many challenges noted in this paper.48 But progress in reducing dependence will remain uneven over the many sectors involved in semiconductor value chains. Chinese companies could make relatively rapid gains in some areas, such as design and OSAT, while facing slower progress in areas such as manufacturing and materials.49 This process will play out over a long time frame, at least a decade or more, with major consequences for the sector, particularly if China succeeds in duplicating major portions of global value and supply chains.


Source : American Affairs


Press Release (April 27,2021): Total Telecom Reports That the Chinese Semiconductor Industry Is Speeding up Advanced Chip Development

Total Telecom finds that China has accelerated its efforts to advance its semiconductor industry in the wake of increasing geopolitical tensions with the US.

China has acquired new capability to produce 28-nm chip wafers, one of several demonstrations of the growing capabilities of Chinese chipmakers, this will likely become mature this year. Total Telecom reports that it is a sign that China is fast developing high-end skills in chip manufacturing.

Chinese companies have started moving their 14nm chipset orders from Taiwan Semiconductor Manufacturing Company (TSMC) to China’s Semiconductor Manufacturing International (SMIC). By the coming year, China is expected to have gained required expertise in the manufacturing of 14nm chips and will also run trial production of 7nm chips this month. They are likely to introduce mass production in October.

Total Telecom says that the geopolitical tensions between the US and China are forcing the country to enhance its capability in chip manufacturing.

Semiconductors are the foundation of our modern digital lives. From being a part of your smartphone or laptop to cars, washing machines and medical devices, semiconductors power our lives. Further, all the upcoming technologies, like the Internet of Things (IoT), Augmented Reality, Virtual Reality need chips to turn the vision into reality.

With the surge in demand driven by the COVID-19 pandemic while under sanction, Total Telecom reports that China is changing the market practice from being the world’s largest importer and consumer of semiconductors to be the manufacturer and less dependent on imports for its chip requirements.

By acquiring expertise in new areas like the manufacturing of 14nm and 7nm chips, Total Telecom believes the Chinese semiconductor industry is all set to grow in prominence and well on its way to becoming self-reliant and disrupting the global semiconductor industry.

The World Relies on One Chip Maker in Taiwan, Leaving Everyone Vulnerable

Yang Jie, Stephanie Yang and Asa Fitch wrote . . . . . . . . .

The company makes almost all of the world’s most sophisticated chips, and many of the simpler ones, too. They’re in billions of products with built-in electronics, including iPhones, personal computers and cars—all without any obvious sign they came from TSMC, which does the manufacturing for better-known companies that design them, like Apple Inc. and Qualcomm Inc.

TSMC has emerged over the past several years as the world’s most important semiconductor company, with enormous influence over the global economy. With a market cap of around $550 billion, it ranks as the world’s 11th most valuable company.

Its dominance leaves the world in a vulnerable position, however. As more technologies require chips of mind-boggling complexity, more are coming from this one company, on an island that’s a focal point of tensions between the U.S. and China, which claims Taiwan as its own.

Analysts say it will be difficult for other manufacturers to catch up in an industry that requires hefty capital investments. And TSMC can’t make enough chips to satisfy everyone—a fact that has become even clearer amid a global shortage, adding to the chaos of supply bottlenecks, higher prices for consumers and furloughed workers, especially in the auto industry.

The situation is similar in some ways to the world’s past reliance on Middle Eastern oil, with any instability on the island threatening to echo across industries. Companies in Taiwan, including smaller makers, generated about 65% of global revenues for outsourced chip manufacturing during the first quarter of this year, according to Taiwan-based semiconductor research firm TrendForce. TSMC generated 56% of the global revenues.

Being dependent on Taiwanese chips “poses a threat to the global economy,” research firm Capital Economics recently wrote.

TSMC, which is listed on the New York Stock Exchange, reported $17.6 billion in profits last year on revenues of about $45.5 billion.

Its technology is so advanced, Capital Economics said, that it now makes around 92% of the world’s most sophisticated chips, which have transistors that are less than one-thousandth the width of a human hair. Samsung Electronics Co. makes the rest. Most of the roughly 1.4 billion smartphone processors world-wide are made by TSMC.

It makes as much as 60% of the less-sophisticated microcontrollers that car makers need as their vehicles become more automated, according to IHS Markit, a consulting firm.

TSMC said it believes its market share for those microcontrollers is about 35%. Company spokeswoman Nina Kao refuted the idea that the world depends too much on the company, given the many areas of specialization in the world’s semiconductor supply chain.

Chipping Away

Over the past few years, TSMC has increased its R&D spending—and saw its market cap become the biggest in semiconductors.

The U.S., Europe and China are scrambling to cut their reliance on Taiwanese chips. While the U.S. still leads the world in chip design and intellectual property with homegrown giants like Intel Corp., Nvidia Corp. and Qualcomm, it now accounts for only 12% of the world’s chip manufacturing, down from 37% in 1990, according to Boston Consulting Group.

President Biden’s infrastructure plan includes $50 billion to help boost domestic chip production. China has made semiconductor independence a major tenet of its national strategic plan. The European Union aims to produce at least 20% of the world’s next-generation chips in 2030 as part of a $150 billion digital industries scheme.

In March, Intel announced a $20 billion investment to build two new chip factories in the U.S. Three months earlier, then-chief executive Bob Swan had flown a private jet to Taiwan to see if TSMC would take over some of the manufacturing for its newest generation of chips, people familiar with the meeting said—a contract potentially worth billions of dollars.

TSMC executives were eager to help but wouldn’t do it on Intel’s terms and disagreed on price, one of the people said. The negotiations still aren’t settled, the person said.

Intel ousted Mr. Swan in January as it tries to recover from missteps that left it potentially reliant on TSMC. Intel’s market cap is around $225 billion, less than half that of TSMC’s.

The Taiwanese maker has also faced calls from the U.S. and Germany to expand supply due to factory closures and lost revenues in the auto industry, which was the first to get hit by the current chip shortage.

A meeting between chip and auto makers facilitated by the Biden administration in May saw some progress but left simmering frustrations, with U.S. auto makers feeling they had yet to get detailed plans on TSMC’s efforts to increase production, said people familiar with the meeting.

TSMC said it has taken unprecedented actions and increased microcontroller production by 60% compared with 2020.

Deep pockets

Analysts say that broader trends in the industry, along with TSMC’s hard-driving culture and deep pockets, will make it hard to create a more diversified semiconductor supply chain anytime soon.

Semiconductors have become so complex and capital-intensive that once a producer falls behind, it’s hard to catch up. Companies can spend billions of dollars and years trying, only to see the technological horizon recede further.

A single semiconductor factory can cost as much as $20 billion. One key manufacturing tool for advanced chip-making that imprints intricate circuit patterns on silicon costs upward of $100 million, requiring multiple planes to deliver.

TSMC’s own expansion plans call for spending $100 billion over the next three years. That’s nearly a quarter of the entire industry’s capital spending, according to semiconductor research firm VLSI Research.

Other countries would need to spend at least $30 billion a year for a minimum of five years “to have any reasonable chance of success” in catching up with TSMC and Samsung, wrote IC Insights, a research firm, in a recent report.

U.S. officials have said they believe the chance of a conflict has grown after an increase in Chinese military activity near Taiwan—an issue that was noted in a public rebuke of China issued by Group of Seven leaders this week. Still, many analysts believe China won’t try to reclaim Taiwan in the near future because the move could disrupt its own supply of chips.

Taiwanese leaders refer to the local chip industry as Taiwan’s “silicon shield,” helping protect it from such conflict. Taiwan’s government has showered subsidies on the local chip industry over the years, analysts say.

TSMC’s Ms. Kao said the company’s success comes from being in the right place at the right time, with the right business model. While Taiwan’s government played a crucial role in its founding investment, she said, the company doesn’t receive subsidies to build facilities.

Taking risks

When Morris Chang founded TSMC in 1987 with the idea that more chip companies would outsource production to fabrication plants, or “fabs,” in Asia, success was far from assured.

Mr. Chang—now 89 years old, with a fondness for playing bridge and reading Shakespeare—spent his early years in mainland China and Hong Kong before moving to the U.S. in 1949 to go to Harvard University and then the Massachusetts Institute of Technology. He spent nearly three decades working in the U.S., spending most of his career at Texas Instruments.

When TSMC was founded, titans like Intel and Texas Instruments took pride in designing, branding and making their own chips. Advanced Micro Devices Inc. founder W.J. “Jerry” Sanders III famously declared: “Real men have fabs.”

With the Taiwanese government providing about half of its initial funding, TSMC gained traction by positioning itself as the Switzerland of semiconductors. Companies like Nvidia and Qualcomm found that by pairing with TSMC, they could focus more on design without the hassle of running their own factories, or worrying about handing their intellectual property to a competitor to manufacture. AMD sold off its fabs and became one of TSMC’s biggest customers, as did other major players, until there were only a few advanced chip makers left.

Each new client that TSMC picked up added to the company’s war chest, enabling it to spend heavily on its manufacturing capabilities. “The power of the model didn’t become evident until they reached very large scale. Once that calculation changed, it changed the name of the game,” said David Yoffie, a Harvard Business School professor and former member of Intel’s board of directors.

TSMC doubled down on R&D, even during the global financial crisis. While other firms were cutting back, Mr. Chang raised TSMC’s capital expenditures for 2009 by 42% to $2.7 billion, upgrading its capabilities in time for the smartphone boom.

A pivotal moment came in 2013, when TSMC began work on mass-producing mobile phone chips for Apple, now its biggest customer. Before that, Samsung—which had its own smartphones—had been the exclusive microprocessor supplier for iPhones.

To fulfill Apple’s first order, TSMC spent $9 billion, with 6,000 people working around the clock to build a fab in Taiwan in a record 11 months. TSMC is now the exclusive supplier for the main processors in iPhones.

When TSMC was trying to develop cutting-edge chips in 2014, it reorganized its research and development team to work 24 hours a day, with 400 engineers handing off work over three shifts, current and former employees say. Some employees dubbed it the “liver buster” plan, because they felt working late harmed their livers.

TSMC also bet big on extreme ultraviolet lithography, or EUV, a technology that used a new type of laser to carve circuitry into microprocessors at thinner widths than previously possible, allowing chips to perform at faster speeds.

Intel was the biggest early investor in EUV, committing more than $4 billion to it in 2012. But it was slower than its main rivals in adopting the technology, and skeptical about whether it would work. Eventually, Intel calculated it was a surer bet to try to improve existing ways of handling lithography.

TSMC worked with ASML Holding NV, the only company now able to produce machines that etch chips with EUV lithography, and vaulted ahead.

Peter Wennink, the Dutch company’s chief executive, said that Mr. Chang took TSMC all-in on their partnership about five years ago with just a few words over tea in his Taiwan office. Mr. Chang retired in 2018.

With EUV, TSMC became one of two companies, with Samsung, to make the most advanced chips with the smallest transistors possible, used in the world’s top smartphones.

Intel is accelerating a shift toward EUV under its new CEO, Pat Gelsinger.

In the crossfire

As TSMC became more dominant, it grew harder to maintain its role as a neutral party in the industry, especially as tensions rose between the U.S. and China, two of its most important markets.

In response to growing U.S. pressure on China, TSMC suspended orders from Huawei Technologies Co., once its largest Chinese customer, last year and committed to building a $12 billion factory in Arizona. The Trump administration promised to help secure $3 billion in incentives, according to two people familiar with the situation, but funding hasn’t been allocated so far.

While TSMC’s Arizona factory will help increase chip production on U.S. soil, it won’t catapult the U.S. to the technological edge. The factory is expected to produce what’s known as 5-nanometer technology chips by the time it’s running in 2024. At that point, the cutting edge is projected to be 3-nanometer technology. Those chips will be made by TSMC in Taiwan.

With microcontrollers for auto makers, TSMC has been privately frustrated by the industry’s insistence that it give priority to its orders, people familiar with the matter said. Auto makers curtailed their own orders last year as the pandemic started. By the time demand snapped back, TSMC had committed capacity elsewhere.

Analysts say TSMC has little incentive to reallocate production. The less lucrative auto chips make up only around 4% of its revenues.

As German auto makers began furloughing workers and slashing production late last year with chip shortages deepening, they lobbied the German government to pressure Taiwan. Germany’s economy minister, Peter Altmaier, wrote a letter to Taiwanese officials urging them to ensure TSMC expanded supply and warning that the chip shortage could derail the global economic recovery.

Mr. Altmaier recently told a meeting of foreign correspondents in Berlin that talks were continuing, but declined to share details.

In May, luxury car maker Audi furloughed around 10,000 workers as it idled production of some of its bestselling models at two factories.

Dimitris Dotis, the Audi brand specialist at Audi Tysons Corner dealership in Virginia, summed up the situation to customers. “Almost all microchips that go into all new vehicles including Audi come from TSMC in Taiwan,” he wrote. “They expect bottlenecks in the supply chain to last through 2022.”


Source : The Wall Street Journal

Meet the One Chipmaker the Entire World Is Now Depending On

Tyler Durden wrote . . . . . . . . .

There is no doubt about it: Taiwan Semiconductor has emerged as the world’s most important chipmaker.

And now, in the midst of a global semiconductor shortage, it is becoming clear exactly how important the company’s dominance has been.

Not only has TSMC made headlines for proposing to expand production into the United States, as we have documented numerous times, but now it is making headlines for how it has become the center of the semiconductor world – and how that can leave the world vulnerable.

TSMC’s chips are in “billion of products”, including iPhones, computers and cars, the Wall Street Journal writes in a new profile of the company. The company has slowly become the world’s 11th most valuable company, with a market cap of about $550 billion. The company reported $17.6 billion in profits last year on revenues of about $45.5 billion. TSMC makes “around 92% of the world’s most sophisticated chips,” the report says.

This has led to the U.S., Europe and China looking to cut their reliance on chips out of the Taiwanese company. But that’s a tough task given its contribution globally. The U.S., for example, only accounts for 12% of the world’s chip manufacturing, down from 37% in 1990.

Analysts aren’t confident of there being a more diversified semiconductor supply chain “anytime soon”. They attribute this to TSMC’s “hard driving culture” and “deep pockets”. The industry has become so complex that once one producer falls behind, it becomes tough to catch up.

And it’s going to be even tougher for competition to catch up if TSMC starts to expand in the U.S. After years of investing in R&D since the company’s founding in 1987, TSMC finally “broke through” when it started to mass produce chips for mobile phones for Apple:

“A pivotal moment came in 2013, when TSMC began work on mass-producing mobile phone chips for Apple, now its biggest customer. Before that, Samsung—which had its own smartphones—had been the exclusive microprocessor supplier for iPhones.

To fulfill Apple’s first order, TSMC spent $9 billion, with 6,000 people working around the clock to build a fab in Taiwan in a record 11 months. TSMC is now the exclusive supplier for the main processors in iPhones.”

Less than a decade later, and barely 30 years after its founding, TSMC is the dominant force in the global semiconductor world.

At the beginning of May, we noted that Taiwan Semiconductor was considering bolstering its production in the U.S., and that President Biden’s Commerce Secretary was urging more domestic production. Now, it looks like TSMC could be within striking distance of a serious U.S. expansion.

We also reported last month that TSMC is “weighing plans to pump tens of billions of dollars more into cutting-edge chip factories in the U.S. state of Arizona than it had previously disclosed”, a Reuters exclusive revealed.

The company had already said it was going to invest $10 billion to $12 billion in Arizona. Now, the company is mulling a more advanced 3 nanometer plant that could cost between $23 billion and $25 billion, sources said. The changes would come over the next 10 to 15 years, as the company builds out its Phoenix campus, the report notes.

The move would put TSMC in direct competition with Intel and Samsung for subsidies from the U.S. government. President Joe Biden has proposed $50 billion in funding for domestic chip manufacturing – a proposal the Senate could act on as soon as this week. Intel has also committed to two new fabs in Arizona and Samsung is planning a $17 billion factory in Austin, Texas.

TSMC CEO C.C. Wei said on a call last month: “But in fact, we have acquired a large piece of land in Arizona to provide flexibility. So further expansion is possible, but we will ramp up to Phase 1 first, then based on the operation efficiency and cost economics and also the customers’ demand, to decide what the next steps we are going to do.”

TSMC has also said that talks in Europe regarding expansion have gone “very poorly”, increasing the likelihood that the chip giant will be focused more on the U.S.

There are no plans for a plant in Europe, a TSMC spokesperson said.


Source : ZeroHedge