The Elevator
Somewhere inside Cory Hall at the University of California, Berkeley, in the early 1980s — the precise date unrecorded, the precise floor unknown — a young woman from Shanghai who had arrived in America speaking almost no English stepped into an elevator and met the man she would marry, build a semiconductor empire with, and eventually lose. The encounter between Weili Dai, then an undergraduate studying computer science, and Sehat Sutardja, a doctoral candidate in electrical engineering who had grown up in Jakarta repairing radios at thirteen, belongs to that category of origin story so improbable it resists embellishment. Two children of Asia's diaspora, both drawn to the same public university on the same hill above the San Francisco Bay, both possessed by complementary obsessions — his for the physics of integrated circuits, hers for the alchemy of turning those circuits into something the market would want. They married in 1985. A decade later, with $1 million in starting capital and a kitchen table for a headquarters, they would co-found Marvell Technology Group, a company that at its peak commanded a market capitalization north of $70 billion and shipped the silicon that moved data through the world's hard drives, networking equipment, and wireless devices.
But the arc of Weili Dai's life is not the simple parabola of immigrant ambition fulfilled. It is something more dissonant: a story about what happens when the company you built becomes the company that doesn't want you anymore — and what you do with the decades that remain. She is the only woman to have co-founded a major global semiconductor company. She is also one of the few founders in Silicon Valley history to have been effectively pushed out of her own creation amid an SEC investigation, an accounting scandal, and a boardroom revolution, then to have re-emerged as a billionaire investor placing bets across the next generation of chip startups. The semiconductor industry, which trades on precision measured in nanometers, has never quite known what to do with Weili Dai — a woman who played semi-professional basketball in Shanghai, who describes herself as a "geek," who built Marvell's customer relationships through sheer force of personal connection in an industry that prefers to hide behind spec sheets.
Her story is a test case for a particular kind of question: Can a founder survive the loss of the thing she founded? And if so, what does survival look like?
By the Numbers
The Weili Dai Portfolio
$1.6BEstimated net worth (2024)
$1MStarting capital for Marvell Technology (1995)
$3.4B+Marvell annual revenue at peak during her tenure
7,000+Marvell employees worldwide at her departure
$237MExpected windfall from Alphawave-Qualcomm deal
96.3MShares held in Alphawave IP Group
36Countries where MeetKai AI platform operates
Shanghai to San Francisco, in Seventeen Years
The facts of Weili Dai's childhood compress into a few dense sentences that carry more weight than their brevity suggests. Born in 1961 in Shanghai — the youngest of three children in a family where the father was a university professor and the mother a teacher — she grew up in a household that treated education as both salvation and obligation. China in the 1960s and 1970s offered brilliant children few paths that did not run through political orthodoxy; the Dai family's intellectual orientation was a form of quiet defiance. The young Weili, tall and athletic, channeled her competitive instincts into basketball, playing at a semi-professional level before she had any notion that her future lay in silicon rather than sport.
In 1978 or 1979 — the sources vary by a year, as immigrant timelines often do — the family emigrated. Dai was seventeen. She arrived in San Francisco speaking almost no English, a detail she would return to again and again in speeches over the following decades, not as a bid for sympathy but as evidence of a thesis: that the distance between where you start and where you arrive is the truest measure of what is possible. The Bay Area's Chinese immigrant community in the late 1970s was a dense ecosystem of first-generation strivers, families who had traded the certainties of the old country for the vertigo of the new one. Dai adapted with the speed of the genuinely determined. She enrolled at UC Berkeley and chose computer science — a field that, in the early 1980s, was still more countercultural than corporate, still closer to the spirit of the Homebrew Computer Club than to the stock-option factories it would become.
She graduated in 1984 with a bachelor's degree in computer science. The diploma would prove less important than the institution — Berkeley's engineering ecosystem was a petri dish of immigrant talent, a place where the future co-founders of companies that didn't yet exist crossed paths in elevators and lecture halls and late-night lab sessions. The Sutardja brothers — Sehat and his younger brother Pantas, both of Indonesian-Chinese origin, both obsessed with electronics since childhood — were products of this same ecosystem. Pantas, who had failed a math class in eighth grade before discovering that his true passion lay in understanding electronics from first principles, would eventually earn three degrees from Berkeley (B.S. '83, M.S. '85, Ph.D. '88 in EECS). Sehat, born around 1961 in Jakarta, had been a certified radio repair technician at thirteen. He arrived at Berkeley via Iowa State and never left the orbit of the university that had claimed him. The brothers' shared obsession with building things — circuits, systems, entire architectures — was the kind of passion that makes marriages and companies possible.
Weili married Sehat in 1985. For the next decade, she worked at Canon Research Center America, doing software development and project management. He worked at Micro Linear Corp. and Integrated Information Technology, accumulating patents — he would eventually hold more than 440 — and refining his understanding of what the market needed but didn't yet know how to ask for. They were, in the taxonomy of Silicon Valley couples, a complementary pair: the engineer who could see the chip, and the operator who could see the customer.
A Kitchen Table and a Million Dollars
The founding mythology of Marvell Technology Group is, like most such mythologies, both true and insufficient. In 1995, Sehat Sutardja, Weili Dai, and Pantas Sutardja started the company from a kitchen table — Sehat's kitchen table, though the story tends to blur whose kitchen it actually was — with approximately $1 million in capital. The name, derived from "Marvel," was chosen to signal ambition: they intended to create technological marvels. What they actually created, at first, was a very specific kind of chip.
The mid-1990s were a hinge moment for the data storage industry. The PC revolution was generating torrents of digital information that needed to be written to and read from hard disk drives at ever-increasing speeds. The controllers that managed this data flow — the silicon brains inside every hard drive — were ripe for disruption. Marvell's founding insight was that you could design a fabless semiconductor company (one that designed chips but outsourced manufacturing) focused on high-speed data interface controllers for hard-disk manufacturers. These chips would dramatically improve storage read-write speeds without sacrificing stability or compatibility. It was a gap in the market that nobody had filled, in part because filling it required the kind of deep mixed-signal design expertise that Sehat Sutardja happened to possess at a world-class level.
The division of labor within the founding team was clear from the beginning, and it would define both Marvell's success and the narrative tensions that followed. Sehat was the visionary technologist, the chip architect, the holder of patents. Pantas served as chief technology officer, the detail-oriented engineer who translated Sehat's visions into manufacturable designs. And Weili? Weili was everything else. She took on product planning, market development, company management, and — most critically — the relationship-building that would turn a startup with a clever chip into a company with a customer base.
If a young girl from Shanghai can come to America knowing almost no English, graduate from Berkeley, raise a beautiful family, build a meaningful career, keep her traditional values and principles and co-found one of the largest semiconductor companies in the world, truly anything is possible in America.
The semiconductor industry in the late 1990s was a world of engineering-driven companies where the sales function was often an afterthought — a necessary evil tolerated by the technologists who considered themselves the real value creators. Dai upended this hierarchy, not by diminishing the engineering, but by insisting that the relationship between chipmaker and customer was itself a form of engineering. She cultivated deep, personal connections with the executives at Marvell's largest customers — the hard-drive manufacturers, the networking companies, the consumer electronics firms. Rumors of her professionalism, as one account puts it, "quickly spread throughout the industry." In a sector where deals were often won on the basis of price and specs alone, Dai introduced a different variable: trust.
The IPO and the Acceleration
Marvell went public in 2000, timing its debut to the last gasp of the dot-com bubble. The company's focus on storage semiconductors — less glamorous than the internet-facing firms that were cratering all around it — proved to be a kind of insurance policy. While the NASDAQ collapsed, Marvell's customers still needed hard drives, and hard drives still needed controllers. The company not only survived but grew.
The years between 2000 and 2015 were Marvell's golden age. Revenue climbed steadily, eventually surpassing $3.4 billion annually. The employee count swelled past 7,000 worldwide. The product portfolio expanded from storage controllers into networking chips, wireless communication semiconductors, and system-on-chip solutions for data centers and enterprise networking. Every major market that required the movement of digital data at scale — and by the 2010s, that meant every major market, full stop — was a potential Marvell customer.
Dai's role expanded in parallel. She held a succession of titles — executive vice president, chief operating officer, general manager of the Communications Business Group, corporate secretary, director of the board — that reflected both her operational breadth and the company's reluctance to pin her down to a single function. She was, in the language of management theory, the "integrator" — the person who bridged the gap between what the engineers built and what the market wanted.
From storage chips to a diversified semiconductor powerhouse.
1995Founded from a kitchen table with $1 million in capital.
2000IPO on NASDAQ amid the dot-com bubble.
2004Sehat, Weili, and Pantas receive Ernst & Young Entrepreneur of the Year award.
2010Revenue exceeds $3 billion; product lines span storage, networking, wireless.
2013Founders receive Dr. Morris Chang Exemplary Leadership Award from Global Semiconductor Alliance.
2015Dai named to GSA board of directors; Forbes ranks her 95th most powerful woman in world.
But the golden age contained, as golden ages always do, the seeds of its own unraveling. Marvell's growth had been fueled in part by a culture of founder-driven autonomy — Sehat's technical vision was essentially unchallenged, and the board functioned more as a ratification body than a check on executive power. The company was incorporated in Bermuda, a structure that afforded certain tax advantages but also insulated the founders from some of the governance pressures that publicly traded American companies typically face. The Sutardja-Dai family's combined shareholdings gave them significant influence over corporate decisions. For years, this arrangement worked beautifully. It also meant that when things went wrong, they went wrong fast.
The Unraveling
The trouble at Marvell arrived in stages, each one compounding the last, like a series of semiconductor defects propagating through a chip design. The first SEC enforcement action came in 2008, targeting expense manipulation — specifically, the backdating of stock options, a practice that was epidemic in Silicon Valley during the mid-2000s but no less damaging for being common. The investigation forced a restatement of financial results and cast a shadow over the founders' stewardship.
But 2008 was prologue. The deeper reckoning came in the years that followed, as Marvell faced a constellation of crises that seemed almost designed to test whether a founder-led semiconductor company could survive its founders' blind spots. A lawsuit from Carnegie Mellon University alleged patent infringement, exposing the company to a potential billion-dollar legal liability. An internal accounting investigation raised serious questions about revenue recognition practices. The company's auditors resigned — an event that, in the world of public company governance, is roughly equivalent to a fire alarm going off in the middle of the night. Revenue and margins declined substantially.
The 2017 proxy statement, written by new CEO Matt Murphy — who had been brought in the previous July to clean house — describes fiscal year 2017 as a year that "began in challenging circumstances, facing a potential billion-dollar legal liability, an accounting investigation, resignation of our auditors, substantial declines in revenue and margins, and serious questions about Marvell's strategic direction." The language is corporate, but the subtext is devastating: the founders' Marvell was broken.
The SEC charged Marvell a second time in 2019, this time for revenue manipulation — specifically, for recognizing revenue prematurely in certain transactions. Revenue recognition fraud represents approximately 60 percent of all SEC enforcement actions for accounting manipulation; it is, as one fraud-detection firm noted, "the easiest accounting manipulation to prove and thus offers the highest chance of successful enforcement action." Marvell had tripped the wire not once but twice.
By 2016, the board had been substantially reconstituted. Sehat Sutardja stepped down as CEO. Weili Dai departed as president and director. Pantas Sutardja left his role as CTO. The family that had built Marvell from a kitchen table was, in the space of a few months, removed from the company entirely. Matt Murphy's arrival marked the beginning of what he called a "transformation" — a word that, in the context of corporate governance, is a euphemism for amputation. The new management team narrowed Marvell's strategic focus to its core markets of storage, networking, and connectivity, rebuilt the board, settled the Carnegie Mellon lawsuit, and addressed the accounting issues. The stock price recovered. Total shareholder return in fiscal 2017 was 71.7 percent.
For the Marvell that emerged from the crisis, the transformation was genuine and largely successful. For Weili Dai, it was something else entirely: the loss of the thing she had spent twenty-one years building, under circumstances that made it impossible to separate her departure from the cloud of scandal that precipitated it. The question of how much responsibility the founders bore for the governance failures — versus how much was attributable to the broader cultural norms of Silicon Valley, where option backdating and aggressive revenue recognition were so common they barely raised eyebrows until they did — has never been cleanly resolved. What is clear is that Dai left Marvell in 2016, moved to Las Vegas with Sehat in 2017, and began the difficult work of constructing a second act.
The Geek and the Connector
To understand Weili Dai's particular gift — and its particular vulnerability — you need to understand the paradox at the center of the semiconductor industry. Chips are among the most complex objects ever manufactured by human beings, designed by people who think in gate counts and process nodes and signal-to-noise ratios. The sales process for these chips involves explaining their capabilities to customers who are themselves highly technical — system architects at Dell, hardware engineers at Western Digital, procurement teams at Samsung. The industry's culture is, accordingly, engineer-centric. The technologist is king. Everyone else is support staff.
Dai never accepted this hierarchy, and her refusal to accept it was both her greatest competitive advantage and the thing that made her legibility in Silicon Valley uncertain. She described herself, repeatedly and without irony, as a "geek" — but the geekiness she claimed was not the narrowly technical kind that the industry valorized. It was a broader enthusiasm for technology as a social force, a conviction that semiconductors were not ends in themselves but means of connection. "High technology is meant to improve the lives of all people on our planet," she said, in a formulation that would sound like boilerplate from anyone who hadn't actually spent two decades hand-delivering that technology to developing countries.
Her role at Marvell was, by design, externally facing. She built the company's relationships with its largest customers. She served as an ambassador between the United States and China at a time when the semiconductor supply chain was becoming increasingly geopolitical. She promoted Marvell's partnership with the One Laptop Per Child program, Nicholas Negroponte's quixotic effort to get cheap computing into the hands of children in the developing world. She championed women in STEM fields, not as a branding exercise but with the intensity of someone who understood — because she had lived it — what it meant to be the only woman in the room.
Each of us is always facing challenges. But my mom always taught me to be positive, giving, and forgiving.
The problem with being the connector in an engineer's kingdom is that the connector's contributions are invisible by nature. Sehat had 440 patents. Pantas had three Berkeley degrees and a deep well of engineering publications. What did Weili have? Relationships. Revenue. A network. Things that matter enormously when the company is growing and are the first things discounted when the company is in crisis. The governance failures that led to Dai's departure were attributed to the founding team collectively, but the narrative that formed around the departures — in SEC filings, in business press coverage, in the industry's whisper network — tended to foreground the technical founders and treat Dai as an appendage. She was Sehat's wife. She was the co-founder who wasn't an engineer. The framing was reductive, and it stuck.
The Architecture of a Second Act
The period between 2016 and 2018 was, by Dai's own account, a time of recalibration. She and Sehat moved to Las Vegas — a city that, for a Silicon Valley semiconductor founder, represented a kind of geographical rebuke, a deliberate step away from the ecosystem that had both made and unmade them. They invested in real estate. They invested in technology startups. They occupied themselves in the manner of wealthy people who have suddenly been relieved of the daily purpose that made them wealthy.
But Dai was not built for leisure. In 2018, she co-founded MeetKai with James Kaplan, a younger technologist who served as CEO. MeetKai was an AI-powered virtual assistant — a company that used artificial intelligence to engage in voice-operated, personalized search conversations with users about recipes, books, games, fitness, weather, and more. The Kai platform launched in 36 countries and operated in more than 15 languages. It was a bet on conversational AI at a time when Siri and Alexa and Google Assistant had already staked out the territory and when the broader public still thought of virtual assistants as marginally useful novelties.
Dai's description of MeetKai revealed something about how she understood technology — not as a product category but as a relationship. "Kai was created to make our lives more beautiful, easier, and efficient, all through voice technology and artificial intelligence," she told Lifewire. The virtual assistant was deliberately designed to be genderless, identifying as "they" and "them." It was intended to retain previous conversations, understand context, and foster what Dai called "real-time discussions." These were not engineering specifications; they were social specifications. Dai was, once again, building the bridge between what the technology could do and what the human on the other side actually needed.
MeetKai also positioned itself as the "Official AI Partner" of the Los Angeles Chargers, a detail that suggests Dai's knack for high-profile partnerships had survived the transition from semiconductors to software. The company was privately funded, and its precise financial performance remains opaque — a common condition for AI startups in the pre-ChatGPT era, when the market for conversational AI was large in theory but diffuse in practice.
The Investor's Eye
The more consequential dimension of Dai's second act was not MeetKai but her transformation into a semiconductor investor. In the years following her departure from Marvell, she deployed capital across a series of startups and established companies in the chip industry, leveraging the network she had built over two decades as Marvell's chief connector. She co-founded Silicon Box, an advanced semiconductor packaging company — a bet on the growing importance of chip packaging as a performance bottleneck in the AI era. She accumulated a 96.3 million-share stake in Alphawave IP Group, a fabless semiconductor company focused on high-speed connectivity technology for data centers and AI infrastructure.
The Alphawave position turned out to be spectacularly well-timed. In 2025, Qualcomm announced a $2.4 billion all-cash deal to acquire Alphawave, a transaction that would deliver Dai an estimated $237 million windfall. The deal, expected to close in 2026, validated not only Dai's investment thesis — that connectivity silicon for AI data centers would become a critical chokepoint in the technology supply chain — but also her ability to identify and back companies before the market caught up to their value.
From a technical engineer to a co-founder, from a business leader to an angel investor — each transformation tracked the rhythm of the semiconductor industry's own evolution. In the 1990s, the value was in designing storage chips. In the 2000s, it was in building diversified semiconductor platforms. In the 2020s, it is in the infrastructure layer that connects AI systems to each other. Dai's portfolio reads like a map of these successive value migrations.
The Bridge Between Two Countries
The geopolitical dimension of Weili Dai's career has always been underappreciated. Born in Shanghai, educated at Berkeley, a co-founder of a company that shipped billions of chips to customers on both sides of the Pacific, she occupied a position that few people in the semiconductor industry could replicate: genuine fluency in both American and Chinese business culture, at a time when the relationship between the two countries was defined increasingly by competition rather than collaboration.
Dai served as what several profiles described as an "ambassador of opportunity" between the United States and China. She worked to bring Marvell's technology to developing countries. She served on the Committee of 100, an alliance of the most influential Americans of Chinese origin, a body that operates in the charged space between cultural diplomacy and political influence. She was, in the idiom of the semiconductor supply chain, a bidirectional link — capable of transmitting and receiving signals in both directions.
This positioning became more fraught as U.S.-China relations deteriorated over the course of the 2010s and 2020s. The semiconductor industry, once a model of globalized collaboration — with chips designed in California, manufactured in Taiwan, packaged in Malaysia, and sold in Shenzhen — became the central theater of great-power competition. Export controls, entity lists, restrictions on AI chip sales to China — the policy apparatus of technological decoupling landed directly on the industry that Dai had spent her career inhabiting. Whether her cross-cultural fluency remains an asset or has become a liability in this new environment is a question that the next decade will answer.
Sutardja Dai Hall, and What Gets Named
At the University of California, Berkeley, on the north side of campus near the intersection of Hearst Avenue and LeRoy Avenue, stands a building called Sutardja Dai Hall. Completed in 2009, it houses the Center for Information Technology Research in the Interest of Society (CITRIS), a hub for engineering research and tech innovation. The building was named for Weili Dai, Sehat Sutardja, and Pantas Sutardja in recognition of their philanthropy — a significant donation whose precise amount has been reported variously but which was substantial enough to warrant the university's most visible form of gratitude.
The building is, in a sense, a fossil record of a particular moment. It was conceived and funded when Marvell was ascendant and the Sutardja-Dai family's reputation was unblemished. It was completed before the SEC investigations, before the boardroom revolution, before the departures. It stands there now as an assertion of continuity — a reminder that the contributions of the founders to Berkeley's engineering ecosystem are not erased by what came after. Dai was the first female commencement speaker in the UC Berkeley College of Engineering's 143-year history, delivering her address in the Greek Theatre on May 12, 2012. Her son Nicholas walked across the stage that same afternoon to embrace his mother. Her older son, Christopher, was already a doctoral student at Berkeley.
The family's relationship with the university is not merely philanthropic; it is constitutive. Berkeley shaped them — gave them the technical education, the professional network, the elevator where two strangers met — and they, in turn, shaped Berkeley, funding the infrastructure where the next generation of engineers would learn. The building outlasts the controversy. That is, perhaps, the point.
Loss and What Follows
On September 18, 2024, Sehat Sutardja died. The man who had held more than 440 patents, who had been named Inventor of the Year and inducted into the Junior Achievement Business Hall of Fame, who had been awarded the Indonesian Diaspora Lifetime Achievement Award and the Dr. Morris Chang Exemplary Leadership Award, who had co-founded a company from his kitchen table and watched it grow to a market capitalization of tens of billions of dollars — and then watched it be taken from him — was gone. He was approximately sixty-three years old.
The details of his death are not, as of this writing, a matter of public record in the way that the details of his life are. What is public is the strange, recursive quality of grief when the person you have lost was also the person with whom you built the defining enterprise of your life. Dai's relationship with Sutardja was not merely romantic or familial; it was structural. They were co-founders in the most literal sense — of a family, of a company, of a philanthropic legacy, of a building that bears both their names. To lose him was to lose the other half of a joint biography that had spanned four decades.
Dai has said little publicly about the loss. She has continued her investment activities. She has continued to advocate for women in STEM, for AI innovation, for the metaverse — a term she has used in interviews with the same enthusiasm that she once brought to semiconductor technology, insisting that AI and the metaverse will combine to create a more "efficient" future. The word "efficient" is revealing. It is an engineer's word, a semiconductor designer's word. It suggests that even grief, for Weili Dai, is something to be processed and converted into productive output.
Graduates are always told to follow your passions. Most of the time I find that people follow their enthusiasms—until they run out. Passion is a very different thing. Following your passion means a willingness to give everything you have every day. Anything you aren't truly committed to—night and day—is not really your passion.
The Only Woman
There is a particular kind of loneliness in being the only one. Weili Dai is the only female co-founder of a major global semiconductor company — not one of a handful, not one of a small cohort, but the only one. The semiconductor industry, which employs hundreds of thousands of people worldwide and generates hundreds of billions of dollars in annual revenue, has produced exactly one woman who was present at the founding of a company that reached the scale of Marvell. This statistic is so stark that it functions less as a data point than as an indictment.
Dai has spent decades drawing attention to this absence. She has been named one of Newsweek's "150 Women Who Shake the World," one of CNN International's "Leading Women Innovators," one of Forbes' "100 Most Powerful Women" (2008), one of Forbes' "America's Richest Self-Made Women." She has been profiled, celebrated, held up as a role model. And the industry has remained, in its foundational layer, essentially unchanged. As of 2024, female founders of VC-scalable startups still cite their gender as a top factor in fundraising difficulty. In 2024, according to PitchBook, teams including female founders raised $38 billion — 27 percent more capital than the prior year, but across 13.1 percent fewer deals. The numbers move, but they do not transform.
What Dai grasped early — and what gives her advocacy its particular edge — is that the problem is not one of pipeline alone. There are women studying computer science, women earning engineering degrees, women entering the semiconductor workforce. The problem is architectural: the industry's power structures, its funding norms, its cultural assumptions about who gets to be a founder and who gets to be support staff, are designed — not consciously, but effectively — to filter women out at each successive stage of the journey from engineer to executive to founder to billionaire. Dai navigated this filter because she was married to a co-founder, because she brought irreplaceable skills to the founding team, because she was, in her own description, willing to give everything she had every day. That her path is essentially unreproducible is the point. The system that produced Weili Dai has not produced another one.
A Chip Off the Next Block
In the mid-2020s, Weili Dai's investment portfolio tells a story about where the semiconductor industry is headed. Silicon Box, which she co-founded, is an advanced semiconductor packaging company — a bet on the thesis that as transistor scaling slows, the critical competitive battleground will shift from making individual chips smaller to making assemblies of chips work together more efficiently. Alphawave, in which she holds nearly 100 million shares, designs the high-speed connectivity IP that links processors to memory to networking equipment inside AI data centers. Both bets target the infrastructure layer that makes artificial intelligence possible — not the models themselves, but the physical substrate on which the models run.
This is a characteristic Dai move: she sees the plumbing before the palace. At Marvell, she understood that hard-drive controllers were not glamorous but were indispensable. In her investment portfolio, she understands that connectivity IP and advanced packaging are not the things that make headlines — those are the AI models, the chatbots, the humanoid robots — but they are the things that make the headlines possible. The $2.4 billion Qualcomm acquisition of Alphawave is a validation of this thesis at scale. The semiconductor industry's center of gravity is shifting from computation to interconnection, from the chip to the system, from the transistor to the package. Dai is positioned at every node in this shift.
Her $237 million windfall from the Alphawave deal — expected to close in 2026 — will further consolidate her position as one of the wealthiest self-made women in American technology. But the money is, in some sense, incidental. What matters more is the pattern: a founder who lost her company, who rebuilt her influence through strategic investment, who leveraged a lifetime of industry relationships to identify value before the market recognized it. In the semiconductor industry, this is called "seeing around the corner." Dai has been doing it for thirty years.
The last image: Sutardja Dai Hall in the late afternoon, the sun coming through the glass facade, graduate students bent over laptops in the research bays, the building humming with the low-frequency vibration of servers and ventilation systems. Weili Dai's name is on the wall. The elevator still runs. The next founder — the one who will look back on this building the way Dai looks back on Cory Hall — is probably inside it right now, riding between floors, about to meet someone.
