aiwiki.page
Text size
Page width
English
English中文
Geography / silicon-valley · Places

Silicon Valley

Silicon Valley is the name for the Santa Clara Valley region in Northern California, a global center of semiconductor, software, and internet industries.

30 keywords13 linked from8 not yet writtenWritten by AI
TransistorIntegrated Circu…Stanford Univers…UrbanizationLabor MarketElectronHuman CapitalSemiconductorSilicon Va…

Silicon Valley is a region in Northern California, at the southern end of the San Francisco Bay, that has become the world's leading center for the design of semiconductors, computer hardware, software, internet services, and, more recently, artificial intelligence. The name is an informal label for the Santa Clara Valley, a place that was once known mainly for orchards and farming. Its importance comes from the cluster of firms, research universities, venture investors, and skilled workers that formed there after the transistor and the integrated circuit were developed, and from the way that cluster kept renewing itself through successive waves of technology. Understanding the valley therefore means understanding how a technical invention, a university, military demand, and an unusual financing and labor culture came together in one place.

This entry traces that development from the agricultural era and the early electronics work at Stanford University, through the postwar semiconductor industry founded by William Shockley and the spin-off firms that followed, to the personal computer, internet, mobile, and AI booms. It explains the industrial structure of the region, including fabless chip design, equipment and supply chains, and venture capital, and it describes the human and social side: the universities and talent flows, the startup culture, and the housing and environmental pressures that come with growth. The final sections cover the current AI-driven concentration of capital and housing demand, the disputes over where the region's boundaries lie, and the question of whether its model can be copied elsewhere.

Geography and the origin of the name

Location and boundaries

Silicon Valley is an informal name for the southern part of the San Francisco Bay Area in Northern California, centred on the Santa Clara Valley. The valley is a long, narrow trough about 90 miles (145 km) long, running south-southeast from the southern tip of San Francisco Bay toward Hollister. The Bay fills much of its northern end, the Santa Cruz Mountains rise to the west, and the Diablo Range lies to the east. The valley floor is a flat alluvial plain in a tectonic depression between the San Andreas Fault and the Hayward and Calaveras faults, so its physical form is a product of long-term crustal movement.

Its edges are much less precise than its core. The heart of the region is Santa Clara County, where Palo Alto, Mountain View, Sunnyvale, Cupertino, Santa Clara and San Jose lie, but the economic footprint has spilled into neighbouring San Mateo and Alameda counties and, in some accounts, Santa Cruz County. No law defines the area. Its extent depends on common usage, industry identity and media convention, which is why estimates of its size and population differ. San Jose, the largest city, is often described as the region's de facto centre.

The origin of the name

The phrase reached print through a journalism series rather than an official designation. Don Hoefler, a reporter and editor in the electronics trade press, published "Silicon Valley, U.S.A." in the weekly Electronic News, beginning with the issue dated January 11, 1971. The series traced the history of the California semiconductor industry from the mid-1950s onward, and the name referred to the concentration of chip makers in the area, which used silicon to build transistors and integrated circuits. The Computer History Museum dates the first published use of the nickname to a January 10, 1971 headline, and some histories credit Hoefler's friend Ralph Vaerst with suggesting the title.

Hoefler did not invent the phrase. A December 1970 news item in Electronic Design already used it in quotation marks, and a May 1970 newspaper advertisement referred to a company helping "production people in Silicon Valley." The term seems to have circulated among local marketing staff and engineers before it was popularised. Before this, the same area was known as the "Valley of Heart's Delight," a reference to its orchards, which is the subject of the agricultural section.

Physical setting and climate

The valley has a Mediterranean climate, with warm and very dry summers and mild, wetter winters. Its position between the Bay and the Pacific moderates temperatures, although the coastal ranges limit the direct maritime influence that San Francisco receives further north on the Peninsula. The flat, well-drained floor and the good soils made the valley easy to farm and later easy to build on, and its sunny, mild weather made year-round outdoor work and expansion comfortable.

The valley's relation to San Francisco also mattered. San Francisco, with its harbour and commercial centre, was the Bay Area's main gateway for shipping, banking and transport, and the valley lay within reach of that network. This access shaped where roads, railways and later highways were built, and it tied the agricultural towns of the south to the wider market. The physical setting did not determine the industries that came later, but it made the valley an attractive place to settle, to farm, and eventually to host large campuses, factories and office parks.

Sources used

Agricultural era and the Mission Santa Clara valley

Spanish colonization reached the valley in 1777, when the Franciscans founded the mission, the eighth in California on the banks of the Guadalupe River. Named for Saint Clare of Assisi, it was the first California mission to carry a woman's name. The valley's name, and the "Santa Clara" name of the county and city, derives from it. The site was unstable. In 1779 the mission had to be abandoned because of flood damage, and an earthquake destroyed it again in 1818. The mission was rebuilt on higher ground, and it was secularized in 1836 and returned to the Catholic church in 1846. In 1851 the Jesuit order took over the remaining buildings and set up a center of higher learning that became Santa Clara University.

Before orchards, the land was used for livestock and crops. The earliest European agriculture in the valley was cattle, followed by wheat cultivation. The mission's role in land use is documented in part through its fertile location; the Bay Area's fertile land and large local population made it one of the most prosperous missions in the system. Sources disagree on when fruit trees were first planted. One account credits the missionary Fathers at Mission Santa Clara and Mission San Jose with the first orchards in the mid-eighteenth century, while a history of the prune industry states that the mission's first orchards weren't planted until the mid-1850s. The two dates should be treated as unresolved.

The commercial orchard economy took shape after the gold-rush era. By 1860 the valley's population had grown to over 11,000, and Californians planted orchards across the valley, growing apricots, cherries, prunes, almonds, walnuts, and other fruit. The prune was central. Le petit prune d'Agen, introduced by Louis Pellier in the 1850s, proved ideal for drying because it kept its flavor and texture. Market access improved sharply after the railroad arrived: the completion of the transcontinental railroad in 1868 opened fruit markets to the East and encouraged processing and shipping in the valley. Canning followed quickly. Dr. James M. Dawson created the first commercial cannery, and canneries soon spread across the valley.

By the early twentieth century the valley was among the largest fruit-producing areas in the country. Around 1919, close to 100,000 acres were planted in fruit trees, and the valley's prunes outnumbered the total United States production. By 1939 San Jose had 18 canneries and 13 dried-fruit packing houses, which made it one of the largest canning centers in the world. The region's nickname, the "Valley of Heart's Delight," reflected how much of its economy depended on this work. Almost everyone who lived in the Santa Clara Valley was involved in fruit production, whether through growing, picking, packing, canning, selling, or supporting the industry. Family labor was part of the system: many young people spent time harvesting prunes on their hands and knees. Canning expanded during the First World War, when the production boost during World War I put the Santa Clara Valley firmly in the lead of the fruit packing industry.

Stanford University was founded on land that had been part of this farm economy. Leland Stanford, a former California governor and U.S. senator, and his wife Jane established the university in 1885 as a memorial to their only child, Leland Stanford Jr., who had died in 1884. The founding endowment included land: a $5 million initial grant that also included 8,180 acres of Palo Alto farmland owned by Leland Stanford. The grant also specified agricultural teaching; the Stanfords directed that the Palo Alto estate maintain a farm for instruction in agriculture. The institution opened on October 1, 1891, and the university is often called "the Farm" to this day. Because the campus was rooted in local land and a regional farm economy from the start, the university's later ties to electronics firms grew out of a setting that was already tied to local industry.

For further context, see Stanford University, urbanization, and labor market.

Early electronics and Stanford's engineering tradition

The electronics industry of the Santa Clara Valley grew out of the vacuum tube. In 1904 John Ambrose Fleming patented a two-electrode valve, and in 1906 Lee de Forest added a control grid to create the triode, which could amplify weak signals. These devices let engineers control the flow of electrons through an evacuated glass envelope, which made radio, telephony, and instrumentation practical. The work gave rise to electronics as a discipline distinct from electrical power engineering, with its own theory, measurement methods, and training. Early radio activity around San Francisco Bay also gave the area a foothold. De Forest worked in Palo Alto for the Federal Telegraph Company, which had been founded there in 1909 and made radio transmitting equipment.

Stanford and Frederick Terman

Stanford University had a modest engineering program in its first decades, and its position changed under Frederick Terman. Terman earned his doctorate at MIT, joined the Stanford faculty in 1925, and became dean of the School of Engineering in the mid-1940s. He treated the school as a source of local industry rather than only a source of graduates for firms elsewhere. Terman encouraged students to start their own companies, and he was willing to invest personally in them. He also helped build ties between faculty and nearby businesses, a practice that later shaped the university's relationship with industry.

Two of his students became the most cited example of this approach. William Hewlett and David Packard both studied electrical engineering under Terman in the late 1930s. In 1939 they founded Hewlett-Packard in a garage behind Packard's house in Palo Alto. Their first product was a precision audio oscillator, an instrument used by sound engineers and laboratories. The company's early sales were modest but built a reputation for high-quality test equipment, and the garage later became a local landmark. Accounts often report that Terman lent the founders a small sum to start, and that a studio purchased several oscillators for its sound work, though the exact details vary between sources.

Early local firms as a pattern

Hewlett-Packard illustrates a pattern that repeated in the valley. A university provided technical knowledge and trained people, a small firm turned that knowledge into a specialized product, and the founders stayed near the campus. Another early case was Varian Associates, founded in 1948 by the Varian brothers, who had worked on the klystron, a microwave amplifier developed in the late 1930s with Stanford researchers. These firms were small and tied to instruments, microwave devices, and radio, but they showed that research done at a university could be commercialized locally. The same logic later supported larger ventures, including the Stanford Industrial Park that Terman helped establish in the early 1950s to lease university land to technology companies, a development covered in the section on the wartime and Cold War economy.

Terman's model rested on the combination of a strong engineering school, local knowledge of the instrument and radio trade, and a willingness among graduates to take risks close to home. Those conditions were not unique to Palo Alto, but the concentration of them there made it easier for later firms to form, hire from the same pool, and learn from one another.

Wartime and Cold War defense demand

The valley's electronics base was shaped by defense work well before the chip era. During World War II, the U.S. wartime science administration drew Stanford electrical engineer Frederick Terman away from his small campus lab to run a large secret effort at Harvard on radar countermeasures. The head of U.S. scientific research, Vannevar Bush, wanted Stanford's Terman to lead 800 researchers on the mission. The work was classified, and it gave Terman direct knowledge of the most advanced electronics then being developed. Its focus on radio, radar, and microwave systems, and on electronic warfare in particular, gave Stanford a technical foundation that the Bay Area firms of the next decades would later draw on.

After the war, Terman returned to Stanford in 1946 as dean of engineering with a plan to make the electrical engineering department a center for microwaves and electronics. He used his military contacts to secure funding for the lab from the Office of Naval Research, the Air Force and the Army Signal Corps. By 1947 the U.S. military was funding half of Stanford's engineering school budget. The Electronics Research Laboratory that took shape around this money concentrated on microwave tubes, radar detection and deception, and the study of the ionosphere. The Korean War pushed the lab further into electronic intelligence and electronic warfare systems. Terman also encouraged researchers to leave the lab and found companies that built microwave tubes and systems for the military, and these firms became early examples of the spin-off pattern that defined the region.

Defense demand also supported a wider network of defense electronics firms in the Bay Area in the 1950s and 1960s. Companies that specialized in countermeasures, search receivers, and signal intelligence were founded or staffed by Stanford alumni and faculty consultants, and many depended on government contracts. A historian of the region's defense industry, Stuart Leslie's work on Stanford is one of the main accounts of how Pentagon-funded research shaped local high-tech sectors, and the defense sector remained important to these firms well into the postwar period.

The Cold War then created a far larger market for the new integrated circuit. In the early 1960s, the U.S. government became the primary customer for integrated circuits, despite their high cost, because advanced electronics were crucial for moon missions and missile guidance. In 1962, NASA's Apollo program decided to use Fairchild's chips in the guidance computers that would take astronauts to the moon. By 1964, Fairchild Semiconductor produced more than 100,000 integrated circuits for the Apollo space program. The military side moved in parallel. In the fall of 1962, Texas Instruments won a contract to develop and produce 13 custom circuits for the Autonetics Minuteman II missile guidance platform. By 1965, the Minuteman program overtook NASA's Apollo procurement as the largest single consumer of ICs.

These contracts mattered mainly because they paid for production learning before civilian markets existed. Chip prices were falling fast under this demand; one account notes that chips costing $120 in 1961 dropped to $15 by the end of 1962. Government buyers accepted high prices for reliability and small size, and their volume gave manufacturers steady revenue while they solved problems of yield and contamination. Fairchild's engineers later said that the military market was very important to the company in these early years, and this experience helped the industry move toward the mass-market products that followed.

Historians disagree about how central defense spending was over the longer run. Military demand clearly shaped the region's early microelectronics; one study of the industry notes that by the 1970s the industry derived only one-tenth of its sales from military contracting. Defense engagement also moved in cycles, as that engagement ebbed during the 1970s with stagflation and post-Vietnam cuts, surged again in the 1980s under Reagan's Cold War defense ramp-up, then dropped once more in the 1990s as the Soviet empire collapsed and the internet blossomed. The debate, then, is less about whether defense money mattered than about how much of the region's later growth it explains.

The transistor and Shockley Semiconductor

The transistor came out of Bell Labs, the research arm of the American telephone system. In December 1947, John Bardeen and Walter Brattain achieved the first transistor action using a point-contact device on a slab of high-purity germanium, while William Shockley led the wider solid-state group. Shockley then went on to invent another type of transistor, the junction transistor, which was built on thin slices of different types of semiconductor materials pressed together. That design was easier to manufacture and more rugged than the point-contact version. Bardeen, Brattain, and Shockley jointly received the 1956 Nobel Prize in Physics for their work. The invention's dispute over credit began inside Bell Labs: Shockley wanted the basic patent in his own name, but Bell decided that the patent would include only the names of Brattain and Bardeen, a decision that left Shockley resentful.

Shockley decided to build his own company in 1956. He left Bell for a variety of reasons, but the location he chose was shaped by personal ties and by the research environment of the Peninsula. His aging mother lived in Palo Alto, and this, together with the nearby presence of Stanford, convinced him to start the lab in Mountain View rather than in Southern California. The money came from Arnold Beckman, whose company provided the funding. The move was made possible by the financial backing of Beckman Instruments, and at that time most semiconductor work still centred on germanium. The lab was contractually tied to Beckman for two years, and its main priority was the automated manufacture of diffused silicon transistors, although Shockley kept some freedom to pursue his own inventions.

Shockley recruited aggressively, but many of his former colleagues on the East Coast refused to relocate, so he hired young Ph.D. graduates in physics, chemistry, metallurgy, and electrical engineering instead. Among them were Robert Noyce, Gordon Moore, Jean Hoerni, and Eugene Kleiner. The team's early work included designing crystal-growth equipment that could produce single-crystal silicon, which was difficult because of silicon's high melting point. The choice of silicon mattered: it became the material that later defined the region's name. Shockley's own research direction shifted in ways that frustrated his staff. He decided to abandon the diffused silicon transistor in favor of an invention of his own, the four-layer diode.

Shockley's management style became the main source of trouble. He was widely regarded as a brilliant physicist but a poor manager. He put employees through psychological tests, intelligence tests, and even lie detector tests, and he posted salaries publicly and recorded phone calls. He believed that staff were leaking trade secrets and sabotaging projects, and he therefore withheld findings from his research staff. By 1957 the lab had about 30 employees, and a group of eight, including Noyce, Moore, and Kleiner, asked for Shockley to be replaced. When the board refused, they resigned. Shockley called them the "traitorous eight." Their departure in 1957 led to the founding of Fairchild Semiconductor, and Shockley Semiconductor never recovered from this departure. Looking back, the episode is important less for the fate of one company than for what it released: a group of experienced engineers who carried silicon expertise and a habit of founding new firms into the next phase of the industry.

Fairchild Semiconductor and the spin-off network

Fairchild Semiconductor grew out of a split at Shockley Semiconductor Laboratory. William Shockley started the laboratory in Mountain View in 1956 and hired some of the brightest graduates in physics, chemistry, metallurgy and electrical engineering to produce semiconductor devices. Many staff found his management authoritarian and unpopular, and his research focus did not prove fruitful. After a demand that Shockley be replaced was refused, eight people decided to leave together. Shockley called the departure a betrayal and said they would never be successful, and the group became known as the "traitorous eight." The members were engineers and scientists Julius Blank, Victor Grinich, Jean Hoerni, Eugene Kleiner, Jay Last, Gordon Moore, Robert Noyce and Sheldon Roberts. They reached an agreement with Sherman Fairchild in August 1957 and formed Fairchild Semiconductor on September 18, 1957. The company operated as a division of Fairchild Camera and Instrument. The parent's capital, reported at about $1.3 million, came with an option to buy the subsidiary later. Each founder had also put in $500 of his own money as a stake in the venture, and when they later sold those stakes back, each received $250,000.

The company's technical base was silicon transistors. Fairchild Semiconductor's first products were silicon-based transistors for military and later industrial applications. Two later advances turned it into a foundation of the integrated circuit industry. Jean Hoerni's planar process covered the wafer surface with a silicon-oxide film, which reduced the contamination that had hurt production, and it solved the reliability issues that had plagued Fairchild's "mesa" transistors. Noyce then saw that components did not need to be cut apart individually. Different components could be created in the same wafer and connected along the surface by a line of conductive metal. Fairchild filed a patent application for the planar process in 1959. Jack Kilby had built a germanium integrated circuit at Texas Instruments in 1958, but the industry preferred Fairchild's invention because the transistors in planar ICs were interconnected by a thin film deposit, whereas Texas Instruments' invention required fine wires. On September 27, 1960, a group under Jay Last created the first operational semiconductor IC at Fairchild. The patent dispute with Texas Instruments was resolved in 1966 by an agreement on cross-licensing.

The spin-off process began almost at once. The first Fairchild spin-off, Rheem Semiconductor in 1959, was followed in quick succession by Jay Last's Amelco Semiconductor, where he was joined by Jean Hoerni and Eugene Kleiner, and by Signetics, the first stand-alone IC company. Kleiner also encouraged a lab technician to start his own company, Electroglas, to build specialized manufacturing equipment, an early example of the supplier firms that later surrounded the chip industry. The most important spin-off was Intel. Founded by Moore and Noyce in 1968, Intel Corporation was the most influential and successful of the first generation of Fairchildren. Other notable offspring included Advanced Micro Devices and National Semiconductor, and AMD was founded by Jerry Sanders. Former employees were collectively called "Fairchildren," and counts of their companies vary: a Silicon Valley genealogy chart published by SEMI included 126 semiconductor companies traced directly to Fairchild in a 1986 printing, while another account gives 125 for the same date.

Several conditions made this pattern self-reinforcing. The company's culture, which mixed youth, diversity, risk-taking, and engineering excellence, encouraged people to leave and start their own ventures. Equity also mattered: employee stock options played an important role in individual motivation, encouraging risk taking and creating, in some cases, substantial personal wealth. Management practices travelled with the people who left. Intel built on the Fairchild management approach, which emphasized meritocracy and openness over hierarchy. Because the new firms drew on the same pool of trained process engineers and shared suppliers, know-how spread quickly through the local labor market rather than staying inside one company.

Fairchild itself declined from the late 1960s. It ran into financial hardship, facing new competition from area startups and falling stock prices as a result. The departures of Noyce and Moore from 1967 onward hurt the company. Ownership changed several times afterward: in 1987 it was sold to National Semiconductor, and National later sold an 80% stake to Sterling Group venture capitalists in 1997. The company's lasting importance lies less in its later fortunes than in the firms and habits it left behind.

Cluster formation and industrial scale

Employment figures show how quickly the valley's electronics base grew in the 1960s and 1970s, although the estimates depend on how each region is bounded. One analysis based on federal business-pattern data puts the valley's workforce at about 60,000 in 1970, behind Route 128's roughly 85,000, but the valley's estimate reaches about 110,000 by 1975, ahead of Route 128's roughly 100,000 Route 128 employed about 85,000 workers in 1970 and Silicon Valley about 60,000. In 1975, Route 128 employed about 100,000 workers–but Silicon Valley about 110,000. A different comparison measures the valley against the whole state of Massachusetts rather than a corridor, and it concludes that by 1980 Silicon Valley employed more electronics workers than the entire state of Massachusetts. Saxenian's comparative study adds that Silicon Valley generated triple the number of new jobs between 1975 and 1990. The crossover date therefore varies by source, but the accounts agree on the direction of the shift.

Density was arguably more distinctive than raw headcount. In 1980 Santa Clara County had 71% of its manufacturing employment in electronics, a figure much higher than other centers of electronics production that emerged in the western United States after World War II. The sector's national weight adds context: by 1989 electronics ranked first among US manufacturing industries, and the automobile industry was roughly a third its size employing more than 2.6 million people.

Spin-off dynamics help explain this concentration. A study of Massachusetts's high-tech growth argues that firms born as offshoots of existing companies or university laboratories tended to stay near their origins, which reinforced spatial clustering the tendency for such firms to start up mainly as spin-offs from other high tech firms or from university laboratories and to remain close to their sources fueled the tendency towards spatial concentration. In practical terms, chip makers, equipment suppliers, and skilled workers could draw on one another within a short distance, which is the logic of agglomeration economies.

The landscape changed nearly as fast as the payroll. Within roughly two decades, the valley's economic base moved from farming and canning toward urban electronics production. The region had earlier been nicknamed the "Valley of Heart's Delight" for its bountiful orchards. Land once planted with fruit now hosts corporate campuses, housing tracts, and highways prune and apricot trees covering large areas of what are now corporate campuses, subdivisions, and expressways. Population growth followed: the county rose from about 642,000 residents in 1960 to about 1.06 million in 1970 Population in the County | 642,315 | 1,064,714, and between 1970 and 1980 San Jose rose in national city rankings from 29th to 18th San Jose was the fastest growing city in the United States between 1970 and 1980. The change was also visible at the neighborhood level. In Cupertino, a former orchard district became more valuable once it lay near high-tech employers land values rose and its location relative to the high-tech industry made it into a desirable location. This kind of urban growth brought new demands for housing and roads that the rest of this entry addresses.

The shift also reshaped the labor market. Organized labor in the valley had taken shape in the 1880s, when the Santa Clara Valley labor movement took off in the 1880s as agribusiness boomed. Huge orchards of prunes, apricots, and other fruit flourished, and alongside them, the canning industry expanded. Electronics employers, by contrast, took a consistently hostile line toward unions from the start From the beginning, high tech workers faced an industry-wide anti-union policy.

Figures for this period come from sources that use different geographic boundaries and different counting methods, so precise comparisons should be treated with caution. Growth in the 1960s and 1970s also overlapped with the defense and space spending discussed elsewhere in this entry, including demand from the Cold War era.

Venture capital and the financing system

Venture capital was not part of the Santa Clara Valley's early chip story. In the mid-1950s, few investors in the area were willing to back young technology companies. The first major financing network grew out of the firms that came from Fairchild Semiconductor. According to fellow venture capitalist Arthur Rock, Kleiner led the eight, obtaining a $1.5 million investment from Sherman Fairchild and taking over the new firm's administrative duties. Chip-making itself spread quickly from there. Between 1966 and 1969, no less than 27 new chip ventures were formed by Fairchild émigrés. Arthur Rock also helped fund Intel in its early years. Two the eight, Noyce and Moore, started Intel Corporation in 1968 and Rock also funded them with $2.5 million.

The model of professional venture investing took a more formal shape in 1972. Thomas Perkins and Eugene Kleiner founded a firm that raised $8 million for the first fund. Its location mattered. In 1972, the firm now known as Kleiner Perkins was the first venture capital firm to open an office on Sand Hill Road. Its results drew followers: the stellar performance of Kleiner Perkins's first $8 million fund quickly attracted other similar firms to the Menlo Park area. Sand Hill Road is a 5.6 mile stretch of roadway that cuts straight through Palo Alto, Menlo Park and Woodside. Today the street is nearly every top Silicon Valley company has been the beneficiary of early funding from firms on Sand Hill Road. The concentration also has a cost: Sand Hill Road is ranked by JLL as the most expensive street for office space in the United States.

Other parts of the financing system were built alongside these firms. Banks with close ties to startups mattered as much as investors. Providing traditional banking services while also funding tech projects, it is considered a backbone of the venture capital industry in the U.S. This dependence became visible in 2023, when numerous VC funds, including major players like Founders Fund, Union Square Ventures and Coatue Management, have advised companies in their portfolios to move their funds out of SVB to avoid the risk of being caught up in the potential failure of the bank. Law firms, accounting practices and other professional service providers also built up around the venture community, though the scale of their role is less documented in the sources consulted here.

Venture finance changed how new firms handled risk. Because early-stage companies are uncertain, VC investments have high rates of failure. Investors typically release money in stages, with each tranche tied to technical or commercial milestones, so that weak projects can be stopped before large sums are committed. Stock options allowed startups to offer employees a share of a possible payoff instead of high cash salaries, which made it easier to attract engineers to firms that might not survive. Together, these arrangements lowered the cost of trying a new idea and accepted many failures in exchange for a few very large successes. The result was a pattern of rapid founding, quick experimentation, and frequent closure that other regions have found hard to replicate.

The system has also produced concerns about concentration of risk. Some investors now put money into vehicles that hold a single company. In tech startup land, it's a type of investment fund that typically involves concentrating all of its assets in one company. Such vehicles have grown as investors clamor to get a piece of hot startups with valuations often in the tens of billions of dollars. Observers have raised questions about whether this structure, along with rising valuations, adds to the bubble risks that have periodically hit the region.

Personal computers, software, and networking

From chips to whole machines

The valley's electronics industry moved from components to complete computers in the 1970s. Apple, founded in 1976 by Steve Jobs, Steve Wozniak, and Ronald Wayne, sold the Apple II in 1977, one of the first personal computers aimed at ordinary buyers rather than hobbyists. The market expanded sharply when IBM's 1981 PC used Intel's 8088 microprocessor. Intel's chips, including the x86 line, became the common hardware foundation for PCs, so a Santa Clara Valley company ended up shaping the architecture of machines built and sold across the world. The industry's ties to microprocessors and personal computers therefore made the region a supplier of standards as well as products.

Much of the interface design that later reached consumers came from Xerox's Palo Alto Research Center (PARC), a lab that developed the Alto workstation with a graphical user interface, a mouse, and early Ethernet networking. Apple engineers and Jobs saw these ideas in 1979, and they shaped the Lisa (1983) and the Macintosh (1984). The episode illustrates a recurring pattern in the region: a large institution produced a research result, and smaller firms commercialized it. Graphical interfaces then pushed software design toward icons, windows, and menus, which in turn raised the importance of operating systems.

Software, databases, and networks

Software became a business in its own right. Oracle, founded in 1977, sold relational database software that let large organizations store and query data with standard commands, and its products became central to corporate computing. Such firms earned money by licensing software and selling support, which gave them recurring revenue that hardware makers rarely enjoyed. Their customers often bought into a single vendor's ecosystem, a dynamic linked to network effects and switching costs.

Networking firms grew as local networks multiplied and needed to be connected. Cisco, founded in 1984 by a Stanford computer group, built routers that moved data between incompatible networks. Through the 1990s it grew rapidly, partly by acquiring small companies with new technology, and it became one of the region's largest hardware suppliers. Cisco's rise connected the valley's older semiconductor expertise to the emerging internet, which was expanding through the World Wide Web.

The dot-com boom and bust

The commercial internet drew heavy investment in the late 1990s. Netscape's browser, released in 1994 by a Mountain View company, helped ordinary users reach the web, and investors began funding internet start-ups in large numbers. Venture capital for the whole country climbed sharply, reaching about $105 billion in 2000 before falling to around $19 billion by 2003. Many start-ups were built to grow first and earn profits later, relying on continuing capital from investors to cover losses. Pets.com, for example, announced its closure in November 2000, less than a year after its initial public offering. The stock market turned on these companies: the Nasdaq fell more than 60 percent from its March 2000 peak during the first quarter of 2001.

The bust removed a large share of the region's firms and jobs. A regional leader later estimated that the collapse eliminated about 150,000 jobs in the valley in a short period, and job losses continued in the following years. Yet the downturn did not erase the region's assets. Many engineers, managers, and founders moved to surviving firms or new start-ups, and the networking hardware, data centers, and fiber built during the boom remained available for later use. Firms that survived, such as Cisco and Oracle, became stable platforms, while the lessons about profitability and creative destruction shaped the more cautious investment practices of the following decade.

Overall, the period from the late 1970s to 2002 shows the valley's characteristic cycle: rapid technological change, abundant financing, sharp contraction, and a residue of skills and infrastructure that supported the next wave of growth. The dot-com episode also tested whether innovation could generate durable profits without a clear business model.

Internet platforms and the mobile era

Google shows how university research in the valley fed into consumer internet services. Two Stanford doctoral students, Larry Page and Sergey Brin, built a ranking method that scored a web page by how many credible sites linked to it, a logic close to how academics weigh citations. Their approach ranked pages by the number of other credible pages pointing to them, rather than by keyword matching alone. The company began in a rented garage in Menlo Park in 1998, and by February 1999 it had moved to an office in the neighboring city of Palo Alto. Its later expansion went well beyond search. Its products came to include email, cloud computing, and the Android operating system. The same pattern of firms growing from local start-ups into global platforms repeated across the region, and Apple, Google, Meta Platforms (Facebook), Netflix, Nvidia, Salesforce, X (Twitter), Uber, and Workday are all headquartered in the area.

The smartphone was the next turning point, and Apple's iPhone was its central example. Steve Jobs presented the device at a San Francisco convention in January 2007, describing it as a touchscreen phone with an iPod, camera, and web-browsing capabilities. It reached U.S. shops in June at a starting price of $499. The first model, however, ran no software written by outside developers. Apple released a software development kit in March 2008, and the App Store opened online on July 10, 2008, with 500 apps available at launch. Developers had to pay a standard $99 annual fee to release apps on the store. The change turned a closed device into a platform that outsiders could extend, and in doing so it created a market for mobile software, from games to social networking and travel tools. Sales of the device reached more than 200 million units by 2012, five years after the iPhone's debut.

Cloud computing moved storage and processing from individual machines into large remote facilities, and the valley became one of the densest places for housing that infrastructure. Northern California, including Silicon Valley, has more than 160 data centers with over 625 megawatts of multi-tenant commissioned power. The market's large data center presence has been attributed partly to the area's history as an internet exchange point, along with its network and cloud ecosystem. The concentration of facilities in Santa Clara has also grown, with one operator's planned campus there expected to provide 72 MW of critical IT load across 540,000 square feet on 12 acres.

These platforms rely on network effects: a service becomes more useful as more people use it, which attracts still more users and makes it hard for newcomers to catch up. This dynamic helps explain why a small number of firms came to dominate search, mobile app distribution, and social networking, and why such concentration has drawn scrutiny from regulators and debate over the handling of user data. The same mechanisms also link the platform era to the data centers that store and process the information these services collect.

Industrial structure: design, equipment, and supply chains

The valley's chip sector is organized around a division of labor. A firm that designs chips without operating any production line is usually called fabless, and such companies design and sell semiconductor chips without owning the manufacturing facilities, known as fabs. The model grew out of a gap in the industry. Integrated device manufacturers had excess production capacity, which gave smaller companies that relied on them a chance to design chips without manufacturing them. The structural turn came with foundries. Morris Chang's chip manufacturer, Taiwan Semiconductor Manufacturing Company, was founded in 1987 as a stand-alone manufacturer. It was dedicated exclusively to manufacturing chips designed by other companies, without competing in design or sales. Once foundries existed, the cost and risk of entering the semiconductor market fell sharply.

Many valley firms that once ran their own plants shifted to this arrangement. Nearly every semiconductor manufacturer once owned its own captive processing plant. Later, production was increasingly farmed out to foundries such as TSMC, UMC, Winbond, and Samsung, and large firms including Qualcomm, Broadcom, and Nvidia adopted the model. The result is a geographic split: the design headquarters stayed in California while fabrication moved overseas. Foundries are at the center of the manufacturing process, and most of them are in Taiwan. TSMC holds roughly 70 percent of the foundry market, which concentrates risk across the industry. A typical example is Nvidia, which designs the most advanced AI accelerator chips but does not manufacture any of them; it sends its designs to TSMC, which fabricates them in fabs in Taiwan, Japan, and Arizona. The arrangement leaves valley firms exposed to geopolitical disruption in the manufacturing base, which is primarily Taiwan.

Design also depends on software. Electronic design automation (EDA) tools provide the software for designing and producing electronic systems, from integrated circuits to printed circuit boards. They let engineers simulate and analyze the performance of circuits before they are physically manufactured. The EDA market is concentrated. It is a duopoly with huge entry barriers, although customers still have considerable bargaining power. Virtually any company designing its own chips uses Synopsys or Cadence, or both. Both firms are major valley-area suppliers, and they reinvest heavily; Synopsys and Cadence consistently devote about 30 percent of revenue to R&D. Designers also license reusable building blocks, so EDA providers usually license certain parts of the fundamental chip design, which saves designers time and money.

Equipment suppliers form another layer. Foundries use a variety of tools, mostly made by five companies: Applied Materials, ASML, Tokyo Electron, KLA, and Lam Research. Applied Materials, headquartered in Santa Clara, supplies tools for patterning, deposition, etching, metrology, inspection, and review of critical steps in forming transistors and interconnects. Lam Research, based in Fremont, supplies wafer-fabrication equipment for etch, deposition, and surface preparation. The region's firms are not present across the whole chain. U.S.-headquartered firms lead in nearly all types of semiconductor manufacturing equipment except photolithography and wafer handling. ASML, a Dutch company, is the sole supplier of photolithography machines used in the most advanced chip manufacturing.

These links make the sector interdependent. Equipment demand follows foundry investment: demand for the whole sector is driven by capital expenditure by foundries such as TSMC, GlobalFoundries, Samsung, and Intel. Problems upstream reach designers first: if foundries face interruptions or delays, fabless companies bear the brunt of potential market setbacks. Hardware, software, and services firms therefore share a dense web of suppliers, licensed IP, and design tools, which is part of why proximity to design and equipment firms mattered even as fabrication moved away.

Universities, talent, and knowledge spillovers

Stanford University and the University of California campuses, especially Berkeley, have been the main engines of the region's technical workforce and research base. Scholars usually credit them with supplying foundational research and trained engineers to local firms. The output is large: in 2016, the institutions in the valley awarded 17,721 science and engineering degrees, about 6,380 more than a decade earlier. The share of those degrees going to women stayed at roughly 37 to 38 percent for years. The region's universities therefore supplied a steady stream of graduates, but the decisive step was how those graduates and their research moved into companies, a process traced in the sections on Stanford's engineering tradition and the spin-off network.

The second channel is the inflow of skilled people from outside the region and outside the country. Immigration figures are the clearest measure. In 1990, roughly one-third of scientists and engineers in the region's technology industries were foreign-born, and most of them came from Asia, especially China and India. By 2000, about 53 percent of the science and engineering workforce was foreign-born, a share well above that in Boston and Austin, where it was under one-fourth. More recent counts from the regional Silicon Valley Index found that more than 60 percent of people working in computer, mathematics, and engineering occupations were foreign-born as of 2020. Skilled immigrants also founded firms at high rates. A study based on companies founded between 1995 and 2000 found that 52.4 percent of the region's engineering and technology companies had at least one immigrant founder. AnnaLee Saxenian described this as a process in which founders bring along former classmates and colleagues from their home countries, who share contacts and market knowledge. She later called the resulting return flow of talent and capital "brain circulation" rather than "brain drain."

The third channel, and the one most often cited by economists, is knowledge spillover. Knowledge is hard to contain inside a single firm. It moves when engineers change jobs, when people meet at conferences and informal gatherings, and when they trade information over informal networks. In Regional Advantage (1994), Saxenian contrasted Silicon Valley with Boston's Route 128. She argued that the valley's firms, often competitors, shared technical knowledge and worked together through a loosely structured network, while the older, more hierarchical and secretive firms along Route 128 did not. Frederick Terman and Stanford helped create the cooperative atmosphere that made this possible. Face-to-face contact seems to have mattered in a measurable way. One study of Silicon Valley found that meetings between workers from different firms were followed by significantly higher mutual patent citations, which the authors read as evidence of local spillovers. Anecdotes point the same way: a Xerox executive recalled in 1991 that about a third of the audience at a PARC seminar were not Xerox employees, yet they joined the discussion actively.

Economists treat these spillovers as a central reason for regional advantage. Knowledge that is partly public can raise the productivity of firms that never paid for it, so markets tend to underinvest in it, which is why spillovers are often used to justify public support for research and innovation. Labor mobility is the mechanism through which much of this happens, and it is reinforced by California's refusal to enforce most post-employment non-compete agreements. Legal scholars such as Ronald Gilson have argued that this legal difference helps explain the mobility and knowledge transfer that Saxenian identified as critical, although the rule was not designed with regional industry in mind. This is discussed in more detail in the section on startup culture.

The effect is not uniform, and the evidence has limits. Research on innovation clusters has found that a large part of Silicon Valley's advantage comes from its connections to other clusters through firms' networks of innovating plants, not only from local spillovers. Saxenian herself has warned in recent years that the largest technology firms have become more inward-looking, which she described as "shutting off the flow of talent," and has argued that the problems she once associated with Route 128 are reappearing in the valley. Whether the region's openness can survive the scale and secrecy of its largest firms is therefore an open question.

Startup culture and wealth

Startup work in the valley is organized around speed and a tolerance for failure. Small teams release early versions of products, measure how users respond, and revise plans in short cycles rather than following multi-year roadmaps. Founders are expected to make large bets with uncertain payoffs, and a failed company is often treated as experience rather than a permanent mark against a career. These norms are reinforced by the venture-financed system described elsewhere in this entry, because investors judge young firms on their potential for rapid growth rather than on steady profits.

Equity is the main instrument that makes this model workable. Young firms usually cannot match the cash salaries of large incumbents, so they offer ownership instead. Employee stock options are described as standard practice in the region and across the United States, and grant sizes are set largely by market benchmarks Employee stock options are standard practice in Silicon Valley and across the US, where grants are driven by the market. Individual stakes are usually small: early employees often receive a fraction of a percent to a few percent of the company, and firms commonly reserve between roughly a tenth and a fifth of total equity in an option pool Startups typically reserve 10% to 20% of total equity for employee compensation through option pools, which is a set of shares reserved for future employees. Early employees might receive 0.5% to 2% equity grants depending on their role, seniority, and joining stage. An option is valuable only if the share price later rises above the price fixed at the grant Stock options give employees the right to purchase company stock at a predetermined price, called the strike price, at a future date. If the company's stock price increases, the employee can sell the stock at a profit.

The structure carries risk for workers. During the 1990s dot-com boom, equity pay spread because cash-poor startups could not afford competitive salaries and used stock to attract talent During the dot-com boom of the 1990s, equity-based compensation became even more prevalent in Silicon Valley. Many start-ups at the time were cash-strapped and could not afford to pay competitive salaries. To attract and retain the best talent, they turned to stocks as a way to offer employees a stake in the company's success. When the bubble burst, many employees were left holding options with no value The dot-com bubble burst in the early 2000s, causing many tech companies to go bankrupt and leaving many employees with worthless stock options. Legal scholars have also noted that many startup employees do not negotiate or closely examine their option agreements, which they describe as a structural weakness of the system Many startup employees do not negotiate or seriously investigate their stock option agreements.

Labor law adds a second mechanism. California has long voided most employee non-compete agreements, and its supreme court has held that such clauses are invalid even when narrowly drawn "[n]oncompetition agreements are invalid under section 16600 in California, even if narrowly drawn." The rule has been in place for more than a century and a half California has effectively banned non-compete agreements for over 150 years, making it the strongest and longest-standing prohibition in the country. Later legislation extended the ban to contracts signed outside the state, so that "any contract that is void under this chapter is unenforceable regardless of where and when the contract was signed." Employers are also barred from attempting to enforce such clauses. Protection for confidential information remains in place, since clauses that prohibit the unlawful or disclosure of company trade secrets remain fully enforceable. The legislature's stated rationale links the ban to economic dynamism: it argued that research shows non-competes "stifle economic development, limit firms' ability to hire[,] and depress innovation and growth." In practice, an engineer can leave for a rival or found a company without a contractual barrier, which moves know-how between firms (see knowledge spillovers).

The gains from this system are unevenly distributed. The 2025 Silicon Valley Index reported that income inequality grew faster in Silicon Valley than in California and the U.S. overall, rising 44% between 2010 and 2023. Wealth is more concentrated still. One CalMatters-reported study found that the top 10% of households holding 66% of the investable assets in the region, and that the top 1% held 48 times more of the total wealth than the bottom 50%, compared to 23 times nationally and globally. A separate index produced by San José State University found that 15% of the affluent region's wealth is held by just nine households. Poverty persists alongside this wealth: about 23% of Silicon Valley residents lived below the poverty threshold in 2021, a 3 percentage point increase from 2019.

Public debate centers on whether this concentration reflects fair rewards or a problem that policy should correct. The venture capitalist Paul Graham has argued that inequality produced by creating companies and jobs is not necessarily harmful, and that anti-poverty efforts should avoid penalizing founders We should attack poverty, he said, but be careful not to go after the people who are creating new companies, jobs, and innovation, just because they are getting rich from doing so. Critics respond that the founder argument does not describe most top earners, since economic studies show that these people are mostly "executives, managers, supervisors, and financial professionals," and not the sort of risk-taking entrepreneurs that Graham is defending. Regional business leaders have defended the system as one with transparent rules that produce winners, while social service providers argue that many families are working full time, often holding multiple jobs, yet still cannot achieve self-sufficiency because wages have not kept pace with the high cost of living. The disagreement therefore concerns not only how much wealth is created, but how its benefits are shared.

Housing, transport, and urban pressures

Housing costs and the jobs–housing gap

Housing has been the most visible cost of the region's employment boom. Median home prices are nearing $2 million, and a quarter of households cannot meet basic needs, according to the 2026 Silicon Valley Index, which pairs record venture investment with this affordability gap. Prices were already high a decade earlier. The 2018 median home price in Silicon Valley was $1.18 million, a 21% increase from 2017. Among the local counties, the median sale price for existing single-family homes was $1.9 million in Santa Clara County, more than double the statewide median and more than four times the U.S. median in September 2024. Since then, the market has stayed near the same level. Santa Clara County's August 2026 single-family median was $1,894,440, down 2.8% from the previous month and about 0.3% below the year before.

The underlying mechanism is a shortage of homes relative to jobs. Between 2010 and 2014, Santa Clara and San Mateo counties gained around 167,000 new jobs but added only 22,000 housing units. The imbalance is sharpest in the job-rich cities. At 3.01 jobs per dwelling, Palo Alto had one of the most unbalanced ratios in the Bay Area, and Palo Alto's median price was pushed to high levels as a result. Rents follow the same pattern. Because employers in the valley compete for workers who cannot easily find affordable homes nearby, housing costs have repeatedly been cited as a leading obstacle to hiring and retention.

Commuting, traffic, and transit

The housing shortage has pushed many workers outward. Middle-class public servants and low-wage workers have moved toward cheaper places, while their jobs remained on the Peninsula, so they spend more time driving. Long commutes have a long history here. A 2018 local account recalled that long commutes were already the norm in the 1960s, when housing costs were a problem. Traffic has grown since then. The typical Silicon Valley auto commuter lost 37 hours to traffic in 2010. A 2014 study also found that employer shuttles can make the imbalance worse, because they let people live farther from work, even as they reduce solo driving.

Public transit developed in stages. The Santa Clara Valley Transportation Authority was created in 1972 under the Santa Clara County Transit District Act. Its light rail began service in 1987, after the agency had started with bus service alone. In 1995 it took on countywide planning and congestion management, when it became the Congestion Management Agency for Santa Clara County. Today the system is large but spread thin. Its network covers a 346-square-mile service area with about 1,236 miles of bus routes and 42 miles of light rail. Commuter rail is run by a separate agency. Caltrain had an average of about 40,000 weekday riders in August 2025, which is modest compared with the number of drivers on the region's freeways. Highways 101 and 280 are the main corridors for commuters, and they are often jammed at peak hours.

Uneven growth and local planning

Growth has not been spread evenly. Job concentration is greatest in affluent cities such as Palo Alto, Mountain View, and Santa Clara, which have far fewer homes than they have workers. San Jose absorbed much of the housing pressure, and it is undergoing extensive construction to meet demand. The city also had one of the largest jobs surpluses in the Bay Area, with 203,576 more jobs than housing units in one 2016 count.

The result has been conflict over land use. In Palo Alto, developers converted lots and small buildings into office space, and residents protested the expansion of commercial development without matching housing. Such disputes have been a recurring theme in local planning, with the question of whether to permit more office construction debated alongside zoning, housing supply, and transit investment. The same tension links the region's economic success to its physical strain: the companies that generate wealth also drive up the cost of the land and the time it takes to get to work.

Environmental legacy and public health

Groundwater overdraft and land subsidence

The valley's environmental legacy began underground. Through the late nineteenth century many wells flowed without pumping, but by the early twentieth century irrigation for orchards and vegetable fields was drawing water out of the groundwater system faster than it could recharge. In the nine years from 1919 to 1928, the water table dropped over 60 feet in the Mountain View-Milpitas area. As the aquifer lost pressure, deep layers of fine sediment compacted and the land surface sank. The problem was first detected in 1933, and by spirit-level surveys in 1967 identified subsidence of substantial magnitude—as much as 8 feet—and areal extent. Downtown San Jose was among the hardest-hit places, where a benchmark marker commemorates roughly 13 feet of ground lowering between 1912 and 1969.

The Santa Clara Valley was the first place in the United States where subsidence caused by groundwater withdrawal was recognized, and it was also the first to take organized remedial action. The water agencies responded with a mix of supply-side and demand-side measures. The Santa Clara Valley Water District began recharge work in the mid-1930s, building dams, importing water, and implementing a pumping tax in 1964. Imported surface water arriving from the late 1960s significantly reduced groundwater pumpage, which was driving the historical subsidence. Agencies also sealed abandoned wells. Subsidence was largely halted by about 1968 to 1969, and artificial recharge represents about one-half of the inflow of water into the valley for the period 1970–1999. The basin remains heavily used, and saltwater intrusion into the shallow aquifer has been held back since the mid-1980s.

Chemical contamination from chip manufacturing

A second legacy came from the chemicals used to make chips. Trichloroethylene, commonly called TCE, was a solvent that chipmakers relied on heavily in the 1960s and 1970s, and it was a powerful solvent used to clean silicon wafers before the chip design was etched onto them. Much of the handling took place at plants where solvents and acids were used in assembly and fabrication, and leaks from underground storage tanks, sumps and pipes were a common source of soil and groundwater pollution. TCE is responsible for nearly half the federal Superfund sites in Santa Clara County.

The Mountain View case illustrates the scale. Chip-makers, including Intel, Raytheon, and Fairchild, ended up leaking thousands of gallons of water contaminated with volatile organic compounds into the soil in Mountain View for two decades. The resulting Middlefield-Ellis-Whisman study area, which the EPA added to the National Priorities List in 1991, includes contamination that can also reach indoor air through a process called "vapor intrusion," a risk that the EPA and local authorities have had to address in buildings above the plume. Contamination also reached drinking water: in 1981, south San Jose residents found they were unwittingly drinking water that contained TCE as well as freon, which had been leaked into the groundwater out of underground storage tanks at IBM and Fairchild Semiconductor. In Sunnyvale, the so-called Triple Site is the legacy of nearly 30 years of chip manufacturing at a single plant. TCE is known to cause cancer, and short-term exposure can cause breathing difficulty and headaches, which is why these sites have drawn public attention. For two months in late 2012 and early 2013, Google employees who worked there were exposed to levels of trichloroethylene (TCE) that exceeded EPA limits when it got into the ventilation system.

Cleanup, regulation and changes in practice

Cleanup has been shaped by federal law. Superfund cleanup rests on the Comprehensive Environmental Response, Compensation, and Liability Act of 1980, as amended in 1986. Site remedies are documented in EPA decision records; for example, the decision for the Teledyne and Spectra-Physics sites in Mountain View describes ground water extraction and soil vapor extraction systems. At the former National Semiconductor site in Santa Clara, the company began in 1982 to close and remove leaking tanks and equipment, installed a pump-and-treat system, and removed soil from selected areas. Neighboring listed sites are also handled through coordinated, area-wide cleanup where their contaminated groundwater overlaps. Remediation continues on many sites today, and some are still operated as pumped-groundwater treatment facilities.

Public health concerns have also pushed the industry toward more rigorous handling of its waste and more transparent monitoring, and the cleanup of groundwater, soil and indoor air has become a standing responsibility of both companies and agencies. The valley's environmental record therefore shows a pattern: a rapid expansion of the manufacturing base in the postwar decades, followed by a long and costly effort to repair the water supply, the ground and the air that industry had used.

The AI boom and renewed concentration

The latest concentration of economic activity in the Bay Area is most visible in venture finance. The nine-county Bay Area drew $100.2 billion in venture capital across 861 deals in the second quarter of 2026, with San Francisco alone capturing nearly $89 billion. Much of that total came from a few very large AI rounds, and one regional report describes the quarter as led by Anthropic's $65B round, as AI powered a record first half. The early months of the year followed the same pattern. In February, AI-related companies comprising 89% of all capital deployed in the month, and the largest deals included Waymo's $16 billion round. Hardware is part of the same wave: investors are backing specialized AI hardware and inference accelerators as a complement to software models. Reports do not agree on Anthropic's figures. One source gives a $30 billion AI round for February, while the Q2 report cites a $65 billion round. The sources do not reconcile these numbers, so they should be checked against the original filings before being relied on.

Office demand has followed the funding. A CompStak and Savills analysis found that San Francisco accounted for 2.9 million square feet of AI leasing in the six-month period, more than in any previous full year tracked by the firms. AI tenants made up 31.3% of new San Francisco leasing in H1 2026, which was close to its 2023 peak, but on much larger volume. Other measures show a bigger share. CBRE reports that AI companies accounted for nearly half of all office leasing by square footage in San Francisco so far in 2026. The overall market has also strengthened: Avison Young counted 12.1M SF leased through the third quarter of 2026, a 49% jump over the same period last year. The demand has been concentrated in particular submarkets. Mission Bay and Showplace Square availability fell from 36.8% to 17.6%. Tenants have also moved from cheap sublease space into longer direct leases; the bargain that drew many early AI tenants is also fading. Analysts remain cautious. CBRE has warned that the trend could reverse as AI begins automating entry-level roles, reducing how much space companies need per employee.

Housing shows the clearest social effect. Redfin's analysis links the price pattern to AI wealth. Since ChatGPT's launch in November 2022, home prices in the Bay Area's luxury ZIP codes with home prices between $3.1 million and $7.6 million saw an average increase of 13.4%, while home values for lower-end properties in the Bay Area—those $535,000 to $615,000—have fallen by 3.8%. The middle tier, from $1.5 million to $2.8 million, rose 6.3%. Overall prices have also climbed: the median home sale price in the San Francisco metro area jumped 14.4% year over year in March to a record $1.7 million. Redfin's economist has described the luxury gains as tied to the very big winners of AI, those executives and those venture capitalists, although the report notes that the link between AI and the luxury housing market isn't strictly causal. Two other forces shape the market. Buyers are often wealthy and pay in cash: such buyers are more likely to pay with cash or other income, rather than a mortgage, and they've benefited greatly from the rising stock market. Real estate brokers have also described a lack of supply, especially of single-family homes, which they call a "mansion shortage". Upcoming share sales by AI startups could add to this wealth; Startups Anthropic and OpenAI are also preparing to go public.

The geography of the boom also matters. Much of the AI-driven leasing and wealth sits in San Francisco rather than in the Santa Clara Valley, which remains the traditional home of chip and hardware firms, including Nvidia's headquarters in Santa Clara. Because the AI boom is centered in San Francisco, the older "Silicon Valley" image of suburban campuses and orchards-turned-offices is no longer a full description of where the region's new money goes. Whether this concentration lasts depends on the durability of AI funding, which analysts say a pullback could quickly reduce.

Boundaries, replication, and scholarly debate

The name has never had a legal definition, and its geographic scope is still disputed. Historically, the concentration has come to be associated with the Santa Clara Valley that lies south of San Francisco. The valley does not map onto a single administrative unit: it is not a readily measurable administrative unit and straddles four counties. Growth has also spilled into neighboring jurisdictions, and the agglomeration has spilled over into adjacent counties such as Santa Cruz County, San Mateo County, and Alameda County. Wikipedia's account describes two uses of the term: a narrower geographic one, referring to Santa Clara County and southeastern San Mateo County, and a metonymical definition referring to high-tech businesses in the entire Bay Area. Because employment, output and venture data are collected by county or metro area, the choice of definition can change the figures a study reports, which is one reason the debate persists.

Saxenian's comparison with Route 128

The most influential comparative study is AnnaLee Saxenian's Regional Advantage (1994), which set Silicon Valley against Route 128 near Boston. The two regions had much in common: both trace their origins to university research and military spending, and both faced severe downturns in the early 1980s. Saxenian's explanation focused on organization rather than technology. Route 128 firms operated by internalization, using policies of secrecy and company loyalty to guard innovation, while Silicon Valley made fuller use of horizontal communication and open labor markets, alongside intense competition among firms. She reports that Silicon Valley generated triple the number of new jobs between 1975 and 1990, a claim that underpins much of her case.

Her argument also challenged the standard economic account of clustering. The literature relies heavily on the concept of external economies to explain the advantages derived from the spatial clustering of economic activity. Saxenian contended that this concept treats each firm as an atomistic unit of production with clearly defined boundaries, and proposed instead to analyze regions as networks of relationships among firms and local institutions. Reviewers generally accepted the contrast but debated its interpretation. Paul Starr located the difference in Silicon Valley's open networks of communication and exchange across firms, and Michael Gertler's critical discussion appeared in Economic Geography in 1995. Because the book was written before the dot-com bust and the later rise of large platform firms, subsequent histories have had to revisit its conclusions, and many of the later accounts of the region's development take up questions of venture finance, talent and firm size that Saxenian treated only briefly.

Imitation and the spread of the name

The name itself has traveled widely. At least 25 places around the world refer to their local high-tech business cluster or industry as Silicon Something. Examples include Silicon Alley in New York, Silicon Glen in Scotland, and Silicon Plateau around Bangalore. Governments have tried to build comparable ecosystems for decades. The Soviet leader Khrushchev was so impressed by what he saw that he led an attempt to replicate it in Akademgorodok in the middle of Siberia, but the planned 'City of Science' failed to produce the Silicon Valley effect. Later political leaders, from Charles de Gaulle to local mayors, tried to emulate the Northern California success story. Today the term functions as a metaphor for innovation as much as a place name: the term "Silicon Valley" has transcended its geographic roots, becoming a metaphor and conceptual symbol for progress and innovation.

How far the model can be copied

Scholars and policy analysts disagree about whether the valley's model can be transferred. One camp stresses that its advantages come from a combination that is hard to assemble on purpose. Research universities, deep venture capital markets, a labor pool drawn from around the world, and a culture that tolerates failure all reinforce one another. A World Economic Forum commentary notes that the economic planners and policymakers who are chasing Silicon Valley's taillights are learning that they cannot always replicate the entrepreneurial culture and finance mechanisms that flourish there. The same commentary argues that the valley's early growth rested on guaranteed demand, which stimulates the most ambitious kind of innovation, mostly from defense and government purchasing, and that this demand side is more replicable than the culture.

Research on cluster policy adds a further caution. The literature on cluster promotion policies finds broad agreement that such policies are unlikely to create clusters from nothing. Some European initiatives have been criticized as artificial cluster initiatives created as a result of entrepreneurial efforts to obtain public funding, rather than reflecting existing regional specialization. The alternative most often proposed is to build on existing strengths in universities, industries and talent rather than copy the valley's labels, a view captured in commentary that urges other regions to innovate around industries and attributes they already have. The debate thus turns on whether the valley is a set of transferable practices, such as specialized suppliers, service providers and firms in related industries, to universities, standards agencies and trade associations, or a historically specific accident of place that cannot be reproduced.

References

  1. Electronic Newsen.wikipedia.org
  2. Silicon Valley, USA, Electronic News - CHM Revolutioncomputerhistory.org
  3. Don C. Hoefler's microelectronics news with manager's casebook - 102714139 - CHMcomputerhistory.org
  4. Don Hoefleren.wikipedia.org
  5. “Silicon Valley” Was Coined by Electronic Design 55 Years Agoelectronicdesign.com
  6. Santa Clara Valleyen.wikipedia.org
  7. Silicon Valley - an overviewsciencedirect.com
  8. Silicon Valleystudy.com
  9. Santa Clara Valley facts for kidskids.kiddle.co
  10. Silicon Valley Region: Geography, Tech Industry, Climate & History Guidegeographyworlds.com
  11. Mission Santa Clara de As%C3%ADsen.wikipedia.org
  12. Brief History of Santa Clara de Asís Missioncaliforniamissions.net
  13. Santa Clara de Asís - California Mission Guidecaliforniamissionguide.com
  14. Sweet Americana Sweethearts: California's Santa Clara Valley-Then and Nowsweetamericanasweethearts.blogspot.com
  15. The Historic Orchardgrpg.org
  16. The Valley of Heart’s Delight - California Prunescaliforniaprunes.org
  17. Cannery Life: Del Monte in the Santa Clara Valley - History San Josehistorysanjose.org
  18. Valley of Heart’s Delightfortheloveofapricots.com
  19. Our Historystanford.edu
  20. Stanford Universityencyclopedia.com
  21. Stanford Universtygoldenbustours.com
  22. History of Stanford Universityen.wikipedia.org
  23. History of Mission Santa Clara de Asíscaliforniamissions.com
  24. Mission Santa Clara de Asisroadsidesecrets.com
  25. Stanford Universitybritannica.com
  26. The Interdependency Of Stanford And Silicon Valleytechcrunch.com
  27. the secret history of silicon valley 7b3steveblank.substack.com
  28. Steve Blank The Secret History of Silicon Valley Part VIII: The Rise of Entrepreneurshipsteveblank.com
  29. Chip War: The Fight for the World's Most Critical Technology - TianPan.cotianpan.co
  30. Early Silicon Valleypbs.org
  31. Minuteman II Missile ICs - 102711695 - CHMcomputerhistory.org
  32. Cold War Armory: Military Contracting in Silicon Valleyfaculty.fiu.edu
  33. Silicon Valley has a long history in the militaryaxios.com
  34. Steve Blank The Secret History of Silicon Valley Part VII: We Fought a War You Never Heard Ofsteveblank.com
  35. From Sputnik to Silicon Valley: The Evolution of the Semiconductor Industrysudalo.medium.com
  36. 1962: Aerospace systems are the first applications for ICs in computerscomputerhistory.org
  37. The Invention of the Transistor: Shockley, Bardeen, and Brattainmicrobasement.com
  38. honoring the trailblazing transistorspectrum.ieee.org
  39. Transistoren.wikipedia.org
  40. William Shockley - Magnet Academynationalmaglab.org
  41. The Traitorous 8 and Birth of Silicon Valleyinvestingcaffeine.com
  42. Shockley’s Historic Semiconductor Laboratory Honored With Two IEEE Milestonesregion8today.ieeer8.org
  43. Shockley Semiconductor Laboratoryen.wikipedia.org
  44. The Traitorous Eight and The Birth of Silicon Valleyeverything-everywhere.com
  45. Traitorous eighten.wikipedia.org
  46. Fairchild Semiconductor - Engineering and Technology History Wikiethw.org
  47. 1956 Nobel Prize in Physicsnokia.com
  48. Fairchild Semiconductoren.wikipedia.org
  49. Fairchild Semiconductorbritannica.com
  50. NIHF Inductee Jean A. Hoerni and the Semiconductor Manufacturing Processinvent.org
  51. Invention of the integrated circuiten.wikipedia.org
  52. Fairchild, Fairchildren, and the Family Tree of Silicon Valley - CHMcomputerhistory.org
  53. Fairchild Semiconductor: Origin of Silicon Valley - LIVE FROM SILICON VALLEYlivefromsiliconvalley.com
  54. The Complete History of Fairchild Semiconductor: Silicon Valley Pioneer - History Toolshistorytools.org
  55. The “Traitorous Eight” and the Rise of Fairchild Semiconductor - Newsallaboutcircuits.com
  56. The Invention of the Integrated Circuit : Jean Hoerni's Patent Notebookthechipletter.substack.com
  57. Fairchild Semiconductor: The 60th Anniversary of a Silicon Valley Legendmedium.com
  58. Silicon Valley and Route 128: The Camelots of Economic Developmentjournal.c2er.org
  59. Origins Of Silicon Valley: The Development of the Electronics Industry in the San Francisco Bay Area - FoundSFfoundsf.org
  60. Regional Advantage: Culture and Competition in Silicon Valley and Route 128 by AnnaLee Saxenian :: SSRNpapers.ssrn.com
  61. Route 128: The development of a regional high technology economy - ScienceDirectsciencedirect.com
  62. About Santa Clara Countyceo.santaclaracounty.gov
  63. The History of Silicon Valley: From Orchards to Microchipstechbloat.com
  64. Economic & Growth Datasantaclaraca.gov
  65. Rancho Rinconada, Cupertino, Californiaen.wikipedia.org
  66. Social Justice Unions Claim Deep Roots in Silicon Valleyreimaginerpe.org
  67. Eugene Kleineren.wikipedia.org
  68. History of Silicon Valleygcase.org
  69. History of Kleiner, Perkins, Caufield & Byersfundinguniverse.com
  70. Sand Hill Roaden.wikipedia.org
  71. A Guide to Sand Hill Road: The Epicenter of Venture Capitalbuiltinsf.com
  72. Newscostar.com
  73. Venture capitalists urge startups to withdraw funds from crisis-laden Silicon Valley Bankcnbc.com
  74. Venture capitalen.wikipedia.org
  75. Silicon Valley bubble risks heighten as investors pile into funds that bet on a single buzzy startupcnbc.com
  76. Kleiner’s Laws8vc.com
  77. Kleiner Perkinsen.wikipedia.org
  78. The First Trillion-Dollar Startuptechcrunch.com
  79. Sand Hill Road is Now the Wall Street of the West Coast • Fresco Capitalfresco.vc
  80. Throwback Thursday: Remembering the Last Big Tech Layoff Bloodbath, the Dot-Com Bust of 2000-2001sfist.com
  81. ‘We’re readjusting’: Silicon Valley ‘firing on most of its cylinders’ despite tech layoffs - Local News Matterslocalnewsmatters.org
  82. The Dot-Com Bubble and Crash (2000-2002): What Happened?marketcrash.pro
  83. 12. Heroes and Villains of the Globalization Era: Dot-Com Boom and Bust: When the Internet Became Big Businesshistoricalconquest.com
  84. 12. Lessons from the Globablization Era: Dot-Com Boom and Bust: When the Internet Became Big Businesshistoricalconquest.com
  85. 1998: How Two Stanford Students Changed the World Foreverretirehub.org
  86. Google is incorporatedhistory.co.uk
  87. Google's Path to Global Dominance Started in This Menlo Park Garage 26 Years Agothenorthlines.com
  88. Northern California Data Centers: Silicon Valleydgtlinfra.com
  89. Steve Jobs debuts the iPhonehistory.com
  90. The revolution Steve Jobs resisted: Apple's App Store marks 10 years of third-party innovationappleinsider.com
  91. App Store launch transforms the iPhone: Today in Apple historycultofmac.com
  92. Silicon Valley Data Centersedgecore.com
  93. IPhoneen.wikipedia.org
  94. Fabless Chip Designers: Shaping the Future of Semiconductorsvaneck.com
  95. Fabless manufacturingen.wikipedia.org
  96. Fabless manufacturing — Grokipediagrokipedia.com
  97. Fabless: The Transformation of the Semiconductor Industrysemiwiki.com
  98. Where did Silicon Valley's Fabs Go?forbes.com
  99. An Overview of the Semiconductor Industrygenerativevalue.com
  100. The Semiconductor Value Chain: Fabless, Foundries, and IDMsibinterviewquestions.com
  101. Semiconductor Business Models: Fabless and Fearless - DigitalVital HUBdigitalvital.org
  102. Understanding the Semiconductor Value Chain - Quartr Insightsquartr.com
  103. The Semiconductor Industry from 10,000 Feet — Eagle Point Capitaleaglepointcap.com
  104. Semiconductors and the Semiconductor Industrycongress.gov
  105. Semiconductor Manufacturing Equipment Companies & Key Players 2032marketsandmarkets.com
  106. Semiconductor Stocks 2026: 60+ Companiesstocktitan.net
  107. Applied Materials (AMAT) - by Sanjiv - Long-term Investingsanj2f3.substack.com
  108. The Strategic Evolution of Fabless Semiconductor Companiesc-suite-strategy.com
  109. Fabless Companies and Business Models-defining and detecting their strength - THE WAVESthe-waves.org
  110. Fabless Design & IP Cores Overviewsemiconductorx.com
  111. Top 10 Semiconductor Equipment Manufacturers & Suppliersverityrank.com
  112. Silicon Valley, Route 128, and Covenants Not to Compete, Thelaw.stanford.edu
  113. Knowledge Spillovers in Silicon Valleynber.org
  114. Innovation Spillovers across U.S. Tech Clusters Xavier Giroud⇤ Ernest Liu†columbia.edu
  115. The knowledge factor in Silicon Valleyamericanhistory.si.edu
  116. (PDF) Silicon Valley in Context: A Comprehensive Analysis of the San Francisco Bay Area's Historical Evolution, Economic Dynamics, and Sustainable Futuresresearchgate.net
  117. Dean Saxenian Sees a Cultural Shift in Silicon Valleyischool.berkeley.edu
  118. Stanford and Silicon Valley - Best Practices in State and Regional Innovation Initiatives - NCBI Bookshelfncbi.nlm.nih.gov
  119. Silicon Valley’s New Immigrant Entrepreneurs • • • AnnaLee Saxenian 1999ppic.org
  120. Silicon Valley's New Immigrant Entrepreneursccis.ucsd.edu
  121. For Silicon Valley, a Worker Pipeline Cut Offvoanews.com
  122. UC Berkeley Newsnewsarchive.berkeley.edu
  123. foreign born engineers dominate bay area tech jobs says joint venture silicon valleyspectrum.ieee.org
  124. Rewarding Talentindexventures.com
  125. Startup Equity Compensation: What All Founders Should Knowbrex.com
  126. Equity Compensation in Silicon Valleyupstock.io
  127. Cable Finaljcl.law.uiowa.edu
  128. The Latest Attack on California’s Ban of Noncompete Agreements - Katz Banks Kumin LLPkatzbanks.com
  129. Non-Compete Laws in California (CA)myworklaws.com
  130. California Extends Prohibition on Noncompete Agreementswsgr.com
  131. California Expands Prohibition Against Non-Competescalemploymentlawupdate.proskauer.com
  132. State of the Valley: Great wealth, flat job growth, and the income…svcf.org
  133. Silicon Valley wealth inequality deepens - CalMatterscalmatters.org
  134. Silicon Valley Pain Index: Nine households hold 15% of wealthkron4.com
  135. income inequality entrepreneursfortune.com
  136. Silicon Valley tech economy is hot, but not everyone benefits - San José Spotlightsanjosespotlight.com
  137. Non-Compete Clause California: 2024 Banusepact.org
  138. Silicon Valley home affordabilityjointventure.org
  139. Affordable housing in Silicon Valleyen.wikipedia.org
  140. The Silicon Valley Real Estate Market Trends Reportpopehandy.rereport.com
  141. In Search of Cheaper Housing, Silicon Valley Workers Face Long Commuteskqed.org
  142. Let Them Commute: The Bay Area’s Housing Crunchurbanland.uli.org
  143. Off Deadline: Looking at Silicon Valley's housing quagmire - Palo Alto Onlinepaloaltoonline.com
  144. Moving Silicon Valley Forwardnonprofithousing.org
  145. Impacts of Silicon Valley Shuttles on Commute & ...escholarship.org
  146. About VTAvta.org
  147. LAFCO OF SANTA CLARA COUNTY SANTA CLARA VALLEY TRANSPORTATION AUTHORITYsantaclaralafco.org
  148. Caltrainen.wikipedia.org
  149. Santa Clara Valley Land Subsidenceexplore.museumca.org
  150. Land Subsidence in the Santa Clara Valleyusgs.gov
  151. Hydrologic framework of the Santa Clara Valley, Californiapubs.geoscienceworld.org
  152. Silicon Valley's Toxic Past Haunts Sunnyvale Neighborhoodkqed.org
  153. The Secret History of Silicon Valley and the Toxic Remnants of the First Computersgizmodo.com
  154. Bay Area tech offices sit on some of the most contaminated land in America, new book claimssfgate.com
  155. SANTA CLARA VALLEY, CALIFORNIApubs.usgs.gov
  156. Santa Clara Valleygroundwaterexchange.org
  157. How Valley Water Resolved Subsidence in Santa Clara Countyvalleywaternews.org
  158. Silicon Valley Superfund Sitesdisasterarea.home.blog
  159. semspub.epa.govsemspub.epa.gov
  160. The Toxic Danger of Silicon Valley Superfund Sites - Air Filters for Clean Aircleanair.camfil.us
  161. Bay Area Venture Funding Surges to $100B in Q2 2026, Powered by Anthropic's $65B AI Round - The Registry Northern California Real Estatenews.theregistrysf.com
  162. Bay Area Tech Funding Climate 2026: a Data-Driven Updatesfbayareatimes.com
  163. Bay Area AI funding February 2026: Key Trendssfbayareatimes.com
  164. San Francisco AI startup funding surge February 2026sfbayareatimes.com
  165. AI Tenants Lease 4.6 Million Square Feet in San Francisco and Manhattancitybiz.co
  166. AI Office Leasing Shifts Toward Direct Class A Dealsfinance.yahoo.com
  167. AI Leasing Race Heats Up Between New York And San Franciscotherealdeal.com
  168. San Francisco Office Market Climbs On AI Leasingbisnow.com
  169. AI Office Leasing Shifts Toward Direct Class A Deals - CRE Dailycredaily.com
  170. AI Office Leasing Hits 22.7%: 2026 CRE Guidetheaiconsultingnetwork.com
  171. Bay Area luxury home prices surge 13% since AI boom, Redfin findsfoxbusiness.com
  172. AI is quietly splitting the housing market in two: Bay Area luxury homes are up 13%, affordable ones are collapsingfortune.com
  173. Luxury Bay Area Homes Surge as AI Economy Widens Housing Gap - The MortgagePointthemortgagepoint.com
  174. AI wealth is creating a 'mansion shortage' and upending San Francisco's housing marketnpr.org
  175. 1000+ San Francisco AI Startups 2026growthlist.co
  176. How AI Wealth Is Driving Silicon Valley Home Prices in 2026strasersiliconvalley.com
  177. Silicon Valleyen.wikipedia.org
  178. Regional Advantage: Culture and Competition in Silicon Valley and Route 128 - Saxenian, AnnaLee: 9780674753396 - AbeBooksabebooks.com
  179. and Industrial Adaptation in Silicon Valley and Route 128huduser.gov
  180. AnnaLee Saxenian > Papers > The Limits of Autarkypeople.ischool.berkeley.edu
  181. Review of Saxenian, <I>Regional Advantage</I>princeton.edu
  182. Other places have tried to replicate Silicon Valley, but they left out one critical componentarchive-share.america.gov
  183. s11365 008 0080 5link.springer.com
  184. All we need is a Silicon Valley: tech place as a strategic branding toollink.springer.com
  185. Can Silicon Valley be replicated?weforum.org
  186. Re-thinking cluster policies: the role of shared vision and Place Leadership on the development of resilient clustersncbi.nlm.nih.gov
  187. Silicon valley cannot be replicatedthenextweb.com
  188. AnnaLee Saxenian: Regional Advantagelink.springer.com