[Apple Shockwave]⑫ EUV Adoption Abandonment 'Butterfly Effect'... The Beginning of US Advanced Semiconductor Retreat

Introduction of ASML Equipment Changes Semiconductor Industry Trends
Intel Provides Advanced Technology and Development Funds but Ultimately Abandons EUV Adoption
The Seeds Intel Planted 30 Years Ago Bloom in Taiwan and Korea

Editor's Note[Apple Shockwave] is a content series that examines the upheaval caused by Apple entering the semiconductor market. You might wonder, why semiconductors from Apple? Apple is no longer just a company that makes smartphones and computers. After long efforts starting from the late founder Steve Jobs, Apple has designed world-class semiconductors used in mobile devices. If Intel was the leader in the PC era, Apple has become the top predator in the mobile era semiconductor ecosystem. Amid the global semiconductor supply chain crisis and large-scale semiconductor production line investments, we will carefully examine the upheaval and prospects in the semiconductor market brought about by Apple Silicon to broaden our readers' insights. Apple Shockwave will visit readers every Saturday. After more than 40 installments, it will be published as a book.

"You want me to invest $200 million in something that might not even be possible?" (Andy Grove, former Intel CEO)

"Yes, this is what research is." (John Carruthers, former Intel R&D head)

Since 2015, Intel has postponed the introduction of 10nm and 7nm processes without explaining the reasons. Most experts agree on one reason: the delay in adopting extreme ultraviolet (EUV) photolithography equipment. This equipment is made only by ASML. Without it, advanced semiconductor chips produced with processes below 10nm cannot be manufactured. It is an essential item that must be secured to compete in advanced processes.

By 2020, half of the EUV equipment sold by the Dutch company ASML was installed at Samsung and TSMC, while Intel just watched. What expression would Andy Grove, who laid the foundation for developing and producing ASML's EUV equipment and is known as the 'super secondary' in the semiconductor industry, have if he knew this?

Andy Grove, former Intel CEO, readily agreed in 1992 to a request from an internal research head for $200 million to invest in EUV. Even John Carruthers, the R&D head who pushed for this decision, was surprised because the strict Grove signed the expenditure approval without hesitation.

Grove, who made the surprising decision to abandon DRAM, which had fallen behind Japanese companies, and focus on central processing units (CPUs), made another tremendous decision looking 30 years ahead.

Carruthers requested investment in EUV, saying that photolithography technology at the time would soon reach its limits. At that time, EUV at the 13.5-nanometer level was considered impossible. Nikon of Japan ignored EUV investment, essentially refusing to invest in something deemed impossible.

Grove, upon receiving what might have seemed a ridiculous investment request, consulted former CEO Gordon Moore.

"Andy, is there any other way?"

Faced with Moore's question?creator of Moore's Law?Grove had no choice but to say "No."

The semiconductor industry is a competition to integrate more transistors into narrower circuit lines. While design is important, without supporting production processes, semiconductor development and quality cannot be maintained. The technology to put over a billion transistors into a chip smaller than a fingernail is almost magical, but there is a technology that performs this magic. Photolithography, the exposure process that draws the designed circuit onto wafers like taking a photograph, was the decisive foundation for semiconductor industry development. This technology was originally developed in the United States. Intel's rapid progress was possible because of its understanding of and investment in photolithography.

Gordon Moore's 'Moore's Law' would have been impossible without advances in photolithography technology. U.S. photolithography equipment companies became the driving force behind the development of the domestic semiconductor industry. However, the situation changed in the 1980s. Japanese companies launched a major offensive centered on DRAM. The U.S. semiconductor industry was on the brink of collapse. Behind this was Japanese photolithography technology.

Extreme ultraviolet photolithography equipment from ASML. It boasts a price tag of up to 500 billion KRW per unit. Photo by ASML website

U.S. Gave Away Advanced Technology but Ultimately Abandoned Adoption

In 1986, the U.S. photolithography company GAC halted new product development and eventually went bankrupt. Another photolithography company, Ultratech, also disappeared into history. Thus, the U.S. lost photolithography, a core technology for semiconductor production.

A dark horse emerged at this time: the Dutch company ASML. Now a 'super secondary' in the semiconductor industry, without ASML's equipment, semiconductor processes below 10nm cannot be completed. Simply put, making the latest iPhones and Galaxy smartphones is impossible without ASML's EUV equipment. ASML's EUV equipment is the only photolithography equipment supporting single-digit nanometer processes.

ASML, which now controls the semiconductor industry, was established by Dutch electronics and semiconductor company Philips together with equipment manufacturer ASM International as a photolithography company.

ASML continued to develop but could not compete with Japanese companies. Japanese companies like Nikon and Canon advanced, and ASML held only about 10% market share. Then a decisive variable appeared: the U.S. decided to fully support ASML. Amid the collapse of domestic photolithography companies, ASML of the Netherlands was chosen as a reliable partner instead of Japanese companies. The U.S. also had the perception that it could not give its technology to Japanese photolithography companies that supported the invasion of the U.S. semiconductor market by their domestic companies.

The first support for EUV commercialization was initiated by Intel. However, Intel could not develop EUV equipment directly. Intel tried to transfer EUV, which originated from U.S. government research labs, to ASML. The government and Congress hesitated and opposed the technology transfer, concerned about giving advanced technology overseas.

U.S. semiconductor companies including Intel even requested the government to allow ASML to acquire the U.S. photolithography company SVG. Former Intel CEO Craig Barrett supported this, stating, "If this acquisition fails, the U.S. EUV equipment acquisition will be delayed."

It took a long time for EUV equipment to be born. In 2012, Intel, Samsung Electronics, and TSMC acquired shares in ASML and funded EUV development.

Despite Intel laying the groundwork for EUV realization, it ultimately gave up on adopting the equipment itself. The birth of advanced process 'Made in America' chips was thus helplessly delayed. Although Intel belatedly sought to secure EUV equipment, Samsung and TSMC had already advanced far ahead.

Samsung Electronics Chairman Lee Jae-yong is discussing cooperation plans with Peter Wennink, CEO of ASML, during his visit to ASML headquarters located in the Netherlands last June. Photo by Samsung Electronics

The Seeds Intel Planted 30 Years Ago Bloomed in Taiwan and Korea

Intel gave up harvesting the seeds sown by the previous CEO 30 years ago. Following Paul Otellini, who rejected the iPhone, Brian Krzanich, who took over management in 2013, left an indelible scar.

Krzanich, Intel's sixth CEO, ignited Intel's downfall. Cutting R&D investment and reducing development personnel were minor issues. The critical mistake was the enormous misjudgment of delaying EUV equipment adoption. This is where Intel, once an unbeatable semiconductor leader, was pushed to a latecomer position. Some even say that his resignation in 2018 due to an internal affair was positive for Intel.

The seeds Grove planted for the future bloomed 30 years later in Taiwan and Korea. As a result, the chips for Apple iPhones used by Americans can no longer be made in the U.S.

The advertising slogan of a major home appliance company that "a moment's choice controls ten years" is still familiar to consumers. In the semiconductor industry, a moment's choice threatens not just ten years but the survival of a company.

Apple's bold move to create its own semiconductors through ARM design and mergers and acquisitions was a masterstroke. The semiconductors designed by Apple, used in all its products, showed performance that overwhelmed competitors with perfect compatibility with its own operating system, delighting consumers. Apple reduced costs, increased sales, and gained enormous profits.

In contrast, Intel showed that it could lose its decades-long throne in the semiconductor field due to a foolish misjudgment. It was not a 'masterstroke' but a 'move cursed even by gods.' In the semiconductor field, it is even harder to recover from mistakes late.