Huang Chongren: He Studied Medicine, Spent His Life Treating Companies Like Patients

30-second overview: Huang Chongren (1949-2026) was the founder of Powerchip Semiconductor and TSMC. He graduated from the Department of Physics at National Taiwan University, earned a PhD in the United States, taught for seven years, and only at age 38 did he resign from his teaching position to work on scanners. In 1994, he entered the DRAM industry. In 2012, Powerchip listed at NT$0.29 per share, carrying about NT$12 billion in debt, and roughly 270,000 shareholders saw their stocks become worthless. He refused bailouts and refused to step down, selling assets, exchanging technology, and reaching out to friends. Nine years later, he brought the company back to the stock market. During the same period, he also used 30 cents per share to buy back his own company's stock. Saving the company and buying stocks operated on the same set of non-liquidation logic.

The day Kingston Technology's founder David Sun walked into Powerchip's office, he was still thinking about how to turn down the offer.

This was how Huang Chongren later described it in an interview. It was Christmas Eve. Powerchip had just been delisted from the stock market, with about NT$12 billion in debt, and the banking consortium was preparing to auction off the factory buildings. What he wanted to discuss was a difficult-to-propose deal: selling the P3 plant's equipment to Kingston, and then having Powerchip do OEM work for Kingston. Equipment exchanged for cash, production lines kept running. He recounted what he said at the time: "David, you've saved many people. With this money, they can all get through the Spring Festival, and it helps preserve some strength for Taiwan's tech industry."¹

At the end of the conversation, David Sun agreed. In July 2013, Kingston purchased the P3 plant's equipment, Powerchip received nearly NT$5 billion, and then took on Kingston's OEM orders.²

This was the same thing Huang Chongren did his entire life. At the edge of a cliff, he didn't liquidate; he sold a piece of assets, traded for some time, and staked everything on a relationship. This logic saved factories and saved many people's year-end bonuses. The same logic allowed him, when the company's stock price fell to 30 cents, to use his own investment company to buy back the stocks. He passed away on July 31, 2026, at the age of 76.³ Looking back at this path, the hardest part was realizing that those two things were actually done by the same person, with the same judgment.

The PhD Who Didn't See Patients, He Sold Scanners

In 1980, Huang Chongren returned to Taiwan with a PhD in medicine from New York's Mount Sinai School of Medicine.⁴ He didn't walk into a hospital; he walked into a classroom: the Department of Physiology at Taipei Medical University, and also his own alma mater, the Department of Physics at National Taiwan University. Later, he became the chairman of the Department of Physiology at Taipei Medical University.⁵

He graduated from the Department of Physics at National Taiwan University in 1971, then went to the United States to study medicine.⁶ He came from a family where it was hard not to pursue an academic path: his grandfather, Huang Chunqing, was a graduate of the imperial examination system of the Qing Dynasty, and his father, Huang, graduated from the medical school of Tokyo Imperial University and returned to Taiwan to teach at the National Taiwan University College of Medicine, becoming a renowned dermatologist in Taipei at the time.⁷ But according to his own accounts to the media, his father early on saw that he was not suited for clinical work, finding him restless and unable to sit still—people who can't sit still can't treat patients.⁸

In 1987, at the age of thirty-eight, he resigned from his academic position and on August 12th founded UMAX Computer (UMAX). The company made scanners and printers, selling them worldwide under its own brand.⁹ A man who had once measured nerve signals in a physiology lab switched careers to sell machines that turned photographs into digital files.

He made a judgment call: Taiwan had the capability to make things cheaper and faster within others' specifications. This judgment was his first gamble in life, one that would eventually let him win—and also let him lose.

Source: National Yang Ming Chiao Tung University, Central News Agency, TSMC major announcements

Jobs Gave Him a Year's Reprieve; the J710 Collided with iMac the Next Year

UMAX was booming fast. In 1995, UMAX was Taiwan's largest scanner manufacturer, holding 13% of the global market, second only to HP.¹⁰ In certain months of 1997, its sales in the United States even briefly overtook HP.

In 1996, UMAX obtained Apple's authorization to produce Mac-compatible computers and launched the SuperMac series.¹¹ For a company that started in peripheral hardware, this was the ticket to becoming a full-fledged computer manufacturer. The fate of this product line hung entirely on a single licensing agreement.

When Jobs returned to Apple in 1997, he revoked all Mac-compatible manufacturers' licenses at once. The version long circulated in the Chinese-speaking world is: UMAX's performance was halved, and it was sued by Apple to the brink of bankruptcy. This version sounds smooth, but the details in the English technical history don't add up. In fact, Jobs specifically left UMAX alone that year. The reason was that UMAX focused on low-end models under $1,000, a price segment where Apple wasn't strong at the time, and UMAX indicated it wanted to expand Mac's territory into East Asia.¹² UMAX was the only one left. As for lawsuits, Apple never sued UMAX from start to finish. The "sued to the brink of bankruptcy" was a plot that emerged as the story grew larger over time.

The following year, UMAX's newly proposed J710 collided with Apple's upcoming iMac G3. Apple decided not to renew the contract. On May 27, 1998, UMAX ceased production of all Mac OS computers, and the Mac OS 8 license expired in July of that year.¹³ The whole affair was more like being kept in school for a year before truly dropping out the second year.

Huang Chongren flew to San Francisco to meet with Jobs. He recalled the other party's words: "Money is not the issue, but I will absolutely not share technology or people with you."¹⁴ Years later, he reflected: "At that time, I was really upset with Jobs, but I knew he would take Apple to new heights." He also mentioned another argument, this time with Bill Gates: Gates said software worth $100,000 in Windows was only sold to you for $100, and he replied, "Bill, 90% of this we don't want. You're forcing it on me. We only need 10%, so please lower the price."¹⁵ His own footnote reads: "In my life doing PC business, I've argued with Jobs and Bill Gates, and it wasn't bad at all."

These two years of being left behind and then cut off taught him one thing: when your entire product line depends on a single customer, once they turn around, you have no way out. But during those two years, he had actually already placed an even bigger bet. In December 1994, Powerchip Semiconductor was established in Hsinchu Science Park, a cross-industry investment by the UMAX Group, led by him as the group chairman, while UMAX was still operating.¹⁶ This time, they were making DRAM (memory chips used for temporary data storage after computers and phones boot up), which was a completely different world from UMAX's scanners: scanners sell design and brand, while DRAM sells identical specifications at the lowest cost, with prices determined by the global market every day.

Forty-Three Letters, Three Thousand Two Hundred and Eighty-Two Days

Powerchip's foundry venture started with a crushing blow. In March 1995, the first 8-inch wafer fab broke ground; in April 1996 it was commissioned, and by October it was in volume production of 16Mb DRAM.¹⁷ The problem was the market didn't wait. When construction began, a single DRAM sold for seven dollars; by the time volume production officially began, the price had already fallen below one dollar.¹⁸ This kind of collapse is endemic to the entire industry, striking every few years. It takes three years to build a fab, but memory prices can turn over in just three months.

His way of surviving was the same as every time that followed: finding connections. At the end of the 1990s, when United Micro Electronics was considering bringing Powerchip into its camp, he went directly to find Morris Chang at TSMC. According to his own recollection later, he said: "If Powerchip joins the UMC camp, UMC's capacity will surpass TSMC, and you'll go from being first to second."¹⁹ This statement comes from his personal recollection, not from meeting minutes. What can be confirmed is that in June 1999, Powerchip signed a strategic alliance cooperation memorandum with the world's advanced and Mitsubishi Electric, avoiding the fate of being swallowed up.²⁰

In October 2002, Powerchip's P1 factory entered volume production. This factory is often written as "Taiwan's first 12-inch wafer fab," but that's incorrect. TSMC's Fab 12 broke ground in December 1999 and began volume production in early 2002, nearly half a year earlier.²¹ Powerchip's real claim to fame requires a qualifier: Taiwan's first 12-inch wafer fab purpose-built for manufacturing advanced memory.

The real issue wasn't about how big the factory was. In 2004, Taiwan's semiconductor R&D spending accounted for only 6% of revenue, while Japan was 21%, Europe and the US 17%, and South Korea 10%. That same year, Taiwanese companies paid over NT$20 billion in technology licensing fees.²² From 2000 to 2009, Taiwan's five major DRAM manufacturers invested a total of NT$904.8 billion, with market share dropping from a high of 23.2% to just 6.3% in the fourth quarter of 2011. This is the least-told story in the semiconductor industry: on the same island, one production line became Taiwan Enterprise: TSMC with flawless ascension, while another built its factory on someone else's process licensing.

Source: National Policy Research Foundation, North American Intellectual Property Report

In November 2006, Powerchip jointly established Ruijin Electronics with Japan's Elpida, building what was at the time the world's largest 12-inch DRAM factory in Hsinchu Science Park. This joint venture is often described as an equal partnership, but in reality Elpida held about 65% of the shares, with Powerchip Group holding only about 20%—a minority stakeholder.²³ Two years later came the financial crisis. DRAM prices collapsed, and Powerchip lost NT$25.5 billion just in the fourth quarter of 2008 alone, with Taiwan's five major DRAM manufacturers losing a combined NT$159.49 billion that year.²⁴ In March 2009, the government attempted to form "Taiwan Memory Company" to consolidate the industry, but Powerchip couldn't join because its technology was already tied to Elpida, and the plan ultimately fell apart.²⁵

In 2010, Apple, through Japan's Renesas, reached out to Powerchip and entrusted it with all 12-inch driver IC foundry orders—one chip per iPhone, three per iPad, totaling 100 million chips.²⁶ Huang Chongren remembers Apple "sending their legal department, very nervous, telling us not to die." At the time general manager Hsieh Tsai-chu added a detail: when bank auditors were going through the books, they suddenly discovered: "Turns out you've been making stuff for Apple?" The repayment pressure eased immediately.²⁷

On February 27, 2012, Elpida filed for bankruptcy protection. Powerchip's immediate public response was "limited direct impact" because there were no accounts receivable from Elpida.²⁸ The counterpunch came a bit later: Elpida's DRAM process licensing disappeared, and a memory manufacturer without its own advanced technology was like a person who had lost their ventilator. In July of that same year, Powerchip announced it would sell its more than 20% stake in Ruijin Electronics to Micron, for a total of NT$96.96 billion.²⁹

On December 11, 2012, Powerchip delisted at NT$0.29 per share, with book debt of approximately NT$120 billion. The banks' conditions were simple: either Huang Chongren personally guaranteed repayment of NT$80 billion, or he stepped down.³⁰ By his own account, he rejected both: "At that time I only had NT$600-700 million; I needed to repay NT$7.1 billion in three months!" "So I wrote letters to 43 bank holding companies, saying I wouldn't seek relief—I wanted to fight to the end, and I'd definitely repay within three months." Three months later, he managed to raise NT$1.11 billion and paid back the money.

✦ "You have to persist, even if it's just 1%, you have to fight to the end. If I don't do it, the company will fall apart."³¹

For the next few years, he did the same thing over and over: sell off one piece, save another. Selling Ruijin's equity to Micron for cash, selling P3 factory equipment to Kingston and then taking back orders, while the company simultaneously shifted from standard memory to wafer foundry services. This path was actually laid out in April 2008, when he spun off the 8-inch factory into Jingsi Electronics. By 2014, debt had dropped from its peak to just over NT$40 billion.³²

📝 Curator's Note
"Preserving the factory and team" is the summary line used in every report, but it doesn't support the inference that "not a single person was laid off." No Chinese or English materials clearly state how many people Powerchip laid off or implemented unpaid leave between 2008 and 2013. The only available data is at the park level: in January 2009, Hsinchu Science Park had 102,077 employees taking unpaid leave, accounting for 77.2% of the park's total workforce, but no source attributes this ratio to Powerchip as a single company.³³ A narrative that treats the company like a patient being rescued is most easily misread here as "he saved every person's job," and that statement currently has no evidence.

In 2019, the company underwent a business split: the wafer manufacturing of three 12-inch factories was transferred to Powerchip Semiconductor Manufacturing (PowerSMI), while Powerchip itself became a holding company. On December 6, 2021, PowerSMI relisted, with an offering price of NT$49.9 and a closing price of NT$75.9 on the first day—a 52.1% increase.³⁴ From the day of delisting, that's 3,282 days. His words: "We truly, truly repaid the NT$12 billion in debt. The company is back."

Source: Global View Monthly Interview, Central News Agency, United Daily News

The Day It Delisted at 0.29 Yuan — Three Months Later, He Bought Back 400 Million Shares at 0.3 Yuan

On December 11, 2012, the day Powerchip [PSMC] delisted, roughly 270,000 shareholders watched their stock turn into wallpaper. Furious investors formed a self-rescue committee and took to the streets.³⁵ Three months later, starting in March 2013, Huang Chongren—through his personal investment companies, Forceful Yuan and Wisdom Li—launched a public tender offer to buy back Powerchip shares at 0.3 yuan per share, acquiring approximately 400 million shares. In 2020, the company carried out a capital reduction and share swap, with each Powerchip share convertible into roughly 0.6 shares of TSMC. His 400 million shares became approximately 240 million TSMC shares. Based on TSMC's closing price on its first day of listing at 75.9 yuan, media estimates suggest he made at least 18 billion yuan.³⁶

This number alone isn't enough to support a redemptive ending, because standing beside it is another question that no one has answered face-on to this day. Even back then, market participants were asking: "If the company's management truly had the ability to turn things around, they should have urged small shareholders not to sell and to hold on to their shares."³⁷ This question is harder to answer than whether the tender offer was legal or not: at the very moment you are more certain than anyone else that this company will survive, do you choose to buy—or do you choose to call on others to hold?

📝 Editor's Note
Rarely mentioned in this debate is a timeline. The 0.3 yuan public tender offer began in March 2013, while the Kingston transaction that truly gave Powerchip breathing room was signed in June 2013, with equipment delivery in July. In other words, at the moment he moved to buy back shares, he did not yet have a finalized rescue deal in hand. He was buying on his own judgment, not on a contract already signed. This timeline makes both interpretations harder to sustain: those who claim he acted only after knowing the outcome must grapple with this sequence, and those who say he was purely gambling must concede that he was gambling on the very same thing that shareholders were abandoning.

Shareholder anxiety did not end with the relisting. Around 2019, Powerchip was renamed Powerchip Innovation Investment Holdings, and the board floated the idea of distributing cash to shareholders through a capital reduction when cash was sufficient. Some small shareholders feared he intended to take the company private. They formed a self-rescue committee and petitioned the Financial Supervisory Commission, where an incumbent chairperson, Gu Li-xiong, had members kneeling on site. Another group of small shareholders launched a "rallying shares" self-protection action, the first phase of which gathered 20% of the shares to seek a buyer.³⁸

These actions stopped at petitioning and share-gathering, never entering the courtroom. The case that did enter the courtroom was an earlier one. Between 2005 and 2006, he was accused of insider trading under the Securities Exchange Act, related to news of Powerchip's strategic alliance with Nanya Technology and wafer factory transactions. This case took two different paths to its conclusion: the criminal portion was acquitted, while the civil portion resulted in a second-instance ruling by the High Court ordering him to compensate 460,3102 yuan on behalf of 297 investors represented by the insurance center.³⁹ Two parallel lawsuits, two different outcomes—both are facts.

On the day of relisting in 2021, he said that the over 320,000 former Powerchip shareholders finally had their answer: their shares had not turned into wallpaper, and the Powerchip self-rescue committee would become a celebration committee.⁴⁰ This statement was true—and yet not entirely so. Some people sold on the day it was priced at 0.29 yuan and never returned to this table.

He Sent Technology to Hefei — Now That Factory Is Competing for the Same Orders as Powerchip

He is one of the few bosses who speaks plainly in front of cameras. At the COMPUTEX press conference on May 12, 2025, a reporter asked him about the Legislative Yuan passing the third reading to add four national holidays, with business communities expressing strong opposition. His response was: "Industry has already had plenty of holidays. Adding 3 or 4 more days won't kill anyone!"⁴¹ That line made a whole generation of young office workers who had never bought Powerchip stock finally get to know him. But when it comes to sending technology to Hefei, he has never spoken about it so bluntly.

Powerchip, which returned in 2021, is the world's sixth-largest and Taiwan's third-largest wafer foundry, specializing in mature and specialty processes. Its main products are panel driver ICs (the chips that make each pixel on a screen light up), CIS image sensors (the chips in mobile phone cameras that turn light into images), and power management ICs.⁴² These chips don't chase the most advanced nanometer numbers; they value stability, low cost, and large volume. Powerchip is no longer a DRAM manufacturer, nor is it on the same track as TSMC's advanced logic processes. It relies precisely on this mature process market — the very market he personally signed a technology transfer into ten years ago.

In May 2015, Hefei Crystal Joint Integrated Circuit (Nexchip) was established, jointly funded by Hefei Municipal Construction Investment Holding Group and Powerchip Group. The same year, Hefei Municipal Government signed three agreements with Powerchip, transferring technology to the latter. Five key technical executives at Nexchip in its early days, including General Manager Cai Huixia, all came from Taiwan's Powerchip system.⁴³ The products Nexchip made highly overlapped with Powerchip's later main products: panel driver ICs and CIS image sensors.

Ten years later, this story has come full circle. Nexchip became China's third-largest wafer foundry. Its expanded panel driver IC production capacity, combined with China's overall mature process industry's price competition, directly pressured Powerchip's main business. In 2024, China's mature process production capacity already accounted for 34% globally, while Taiwan accounted for 43%, with the gap narrowing. It is generally expected to be overtaken around 2027.⁴⁴ Taiwan's financial and business media titled this story "Helping China Build Nexchip, Powerchip Strikes Itself." Powerchip's original holdings in Nexchip were also diluted through round after round of capital increases by Hefei's local state-owned assets, eventually being sold off in batches.

The pressure landed in Miaoli County's Cuilong. Powerchip's P5 factory was completed in May 2024, with a planned total investment of NT$27.8 billion. By then, about NT$8 billion had already been invested, but with mature process supply exceeding demand, this factory's monthly production capacity was only opened to about 8,000 wafers, far below the 15,000 to 20,000 wafers profit threshold.⁴⁵

On January 17, 2026, Powerchip announced the sale of the Cuilong P5 factory to Micron for US$1.8 billion (approximately NT$56.9 billion). The transaction was completed on March 16, simultaneously launching a technical cooperation with Micron on HBM backend wafer foundry services in the Hsinchu factory area (HBM is the high-bandwidth memory that AI servers are competing for, stacking multiple memory chips to allow faster data input and output).⁴⁶ This was his final major deal in life: selling off his largest bet to regain cash and secure a ticket to a new track. The existing employees at the P5 factory went along with the factory, with additional placements arranged, though neither Micron nor Powerchip said much about it.

May 2, 2024, the completion and opening ceremony of Powerchip Semiconductor's Cuilong Factory. This 12-inch factory, with a planned total investment of NT$27.8 billion, was sold to Micron in less than two years due to price wars in the mature process market. The image does not confirm Huang Chongren's presence.

His public assessment of the Chinese market also changed in his later years. He said Taiwan's semiconductor industry "must withdraw" from the Chinese market, otherwise it cannot survive, because the kind of chip business that needs to be used in China, Taiwanese manufacturers cannot do it.⁴⁷ His technology transfer raised a competitor, and as that competitor grew, it changed the price structure of the very market he was in. His final prescription was withdrawal. This technology transfer itself has not triggered any government's public investigation to date.

At the same time, there was another cooperation that didn't come to fruition. In July 2023, Powerchip signed a cooperation memorandum with Japan's SBI Holdings, planning to build a 12-inch wafer factory in Japan to seek official subsidies. In October, they signed a basic agreement with Miyagi Prefecture Governor Murai Yoshiaki, selecting Sendai as the location, with Japan's Ministry of Economy, Trade and Industry expected to provide subsidies of approximately ¥140 billion. In September 2024, Powerchip notified SBI that it could not proceed with the plan, and the cooperation fell apart. On October 3, SBI President Bei Ji-yoshi quoted the Analects "Without trust, there can be no governance" on Facebook to criticize Powerchip.⁴⁸

Powerchip's explanation was: those were non-binding agreements. Japan's Ministry of Economy, Trade and Industry required the subsidized party to guarantee that the new factory would maintain continuous production for more than ten years, while SBI itself had no semiconductor experience, meaning Powerchip would have to bear a ten-year joint guarantee for a factory it did not control or hold a majority stake in, which would violate Taiwan's Securities Exchange Act. The core of the dispute was a commercial and legal determination of who should bear the joint guarantee responsibility, which eventually evolved into a public opinion battle between both parties. Bei Ji-yoshi's accusations were individual and corporate unilateral statements, not judicial or regulatory agency determinations.

When it comes to gambling on relationships, even those who have won many times can lose. This time, the bill was paid by his own company.

He Said Those Paintings Won't Be Sold — They'll Stay in Taiwan

He rarely appears in obituaries, yet he was once listed among the world’s top 200 leading collectors by ARTnews magazine in the United States.⁴⁹

His collection spans a wide range. He once acquired Portrait of Giovanna for $1.6 million, which sparked his study of Leonardo da Vinci. He also owns items once used by the Qianlong Emperor, among which the most un-collector-like piece is the Heart Sutra that Qianlong reviewed daily and held in his hands — four out of five volumes are in his possession.⁵⁰

His criteria for selecting pieces have nothing to do with investment: "Everything I collect, I like and admire. Even a painting with great potential, if I don’t like it, I wouldn’t even want to buy it." When asked whether he would build a museum like Johnson Lin of BenQ, he said: "Maybe one day, I’ll have the museum exhibit my collection, or I’ll just build my own museum." Then he added a clearer statement:

✦ "I believe I won’t sell my artworks — they’ll stay in Taiwan, for everyone to see."⁵¹

The same person who sold off assets piece by piece to preserve production lines during the company’s darkest hour refused to sell a batch of paintings that could easily be monetized. In his eyes, these paintings were merely temporarily in his possession — meant to remain entirely in Taiwan.

December 20, 2021 — President attends the ribbon-cutting ceremony for the inauguration of the Industry-Academia Innovation Research College. The establishment of these semiconductor-focused research colleges in Taiwan was credited in National Taiwan University’s memorial statement to Huang Chong-jen’s initiative and Powerchip’s leading donation.

He also left behind something even harder to value. In the memorial statement issued by National Taiwan University after his passing, it was written: In 2021, he actively advocated with the government to relax regulations based on industrial needs, and personally reported to then-President Tsai Ing-wen on the urgency of establishing a semiconductor college. Under his call and Powerchip’s leading commitment to long-term donations, it was achieved that National Taiwan University and other top universities in the country established specialized semiconductor research colleges.⁵² This is more concrete than any donation amount. His self-founded Powerchip Culture Foundation followed the same direction. In 2021, it was awarded the Ministry of Culture’s Cultural Contribution Award, and NTU’s introduction described it as “dedicated to preserving world-famous paintings in Taiwan.”⁵³

His inventory list has always been missing one item: As long as something in his eyes “still has the potential to grow,” he wouldn’t allow himself to liquidate it and exit. This applied to the company, the collection, and talent alike. In an interview in 2020, he summed it up: “I’ve seen it all in my life — storms and all. I’ve done medicine, been a professor, worked in high-tech, and collected art. Life has had no complaints.”⁵⁴

"Dividend Will Definitely Be Paid Next Year" — Then Only Seventeen Days Remained

At the April 10, 2026 shareholders' meeting, due to health reasons, he was absent and Deputy Chairman Hsieh Tsai-chu took his place to address shareholders, conveying his message that he would definitely attend in person next year.⁵⁵ On July 14, he attended the institutional investor briefing in person. Someone asked about dividends, and he replied: "Dividends will definitely be paid next year; the specific amount will depend on actual operational performance and the board's resolution."⁵⁶ That was the last time he stood before shareholders in a public setting.

The Monday before his death, he had dinner with Wang Bi-sheng, the then-deputy minister of health and welfare. Wang Bi-sheng was his student when he took his class at National Taiwan University Hospital in 1990. As the dinner was about to end, Wang Bi-sheng originally wanted to take a photo with him, thinking to himself, "Next time."⁵⁷

On the afternoon of July 31, he passed away in his sleep at home due to heart and lung failure, at the age of 76. TSMC announced the major news at 5:51 PM that evening, with Deputy Chairman Hsieh Tsai-chu assuming the duties of chairman by law, and the company also declared that production and sales operations would proceed in an orderly manner according to the established plan.⁵⁸ Three days later, on August 3, TSMC's stock opened at NT$55.2, hovering near flat in the morning session with over 17,000 transactions, relatively calm on the surface.⁵⁹

Financial media gave him the nickname "Nine-Lives Strange Cat." Nicknames for Taiwanese entrepreneurs usually speak to achievements — "Business God" or "Semiconductor Patriarch" — but his nickname spoke of not dying. This name had two layers. In 2002, he was diagnosed with stage II colorectal cancer and recovered after treatment. Ten years later, the company was delisted with NT$12 billion in debt; he didn't run away or declare bankruptcy, and nine years later he brought the company back to the stock market.⁶⁰

He placed his final bet on the earliest of his professions. In an interview in February 2026, he spoke about a biotech new drug he had invested in: a PTHrP-targeted monoclonal antibody aimed at pancreatic cancer, lung cancer, and esophageal cancer. He said: "I am the first person in the world to think about treating cancer by inhibiting the growth of cancer cells; no one has ever thought of it that way."⁶¹ A man who gave up medicine at 38 was talking at 76 about how to make cancer patients suffer less.

Source: ETtoday

Financial media figure Hsieh Chin-ko wrote after his passing that he was "a natural-born optimist," and also said that during their last meeting, the other party seemed to have come specifically to say goodbye.⁶² The word "optimistic" doesn't quite fit him. Optimistic people believe things will get better, so they can wait. What he believed in was that there was still time, so every time he had to call David Sun that night, write to forty-three banks, and push the equipment onto trucks.

On Christmas Eve, he told David Sun that with this money, many people could celebrate the Lunar New Year. More than ten years later, he was still talking about the same thing, except this time the people he was trying to save, he didn't know any of them. The previous transaction he waited for the outcome, this one he didn't. He spent his whole life gambling on "there's still time to finish the job," and the last time, time left first.

Further Reading:

• Semiconductor Industry — A comprehensive map of Taiwan's chip industry, with Powerchip's cliff-edge road being one of the least told stories.
• TSMC — Another flawless uphill climb on the same island; only by comparing the two can you see how unforgiving this industry is.
• Morris Chang — The man Huang Chong-ren sought out before being acquired by United Microelectronics.
• Stan Shih — His Acer partnered with Texas Instruments in 1989 to establish United Microelectronics, Taiwan's first DRAM factory, five years earlier than Powerchip.
• Tsao Hsing-cheng — Founder of United Microelectronics, who had intentions of incorporating Powerchip into his camp in the late 1990s.

Image Sources

• Cover image / Huang Chong-ren receiving an honorary doctorate in engineering (2022): Released from the official announcement page of the Secretariat, National Yang Ming Chiao Tung University, during the awarding ceremony on March 28, 2022. Fair use for editorial commentary. Source: National Yang Ming Chiao Tung University Secretariat.
• Inauguration of the Industry-Academia Innovation Research College (2021-12-20): Presidential Office photograph by Makoto Lin, CC BY 2.0. Source: Wikimedia Commons.
• Completion and opening ceremony of the Powerchip Electronics Copper Locomotive Plant (2024-05-02): Presidential Office photograph by Wang Yu Ching, CC BY 2.0. Source: Wikimedia Commons.

References