Cypress Semiconductor Corporation Business Information, Profile, and History

Company Perspectives:

CYPRESS CORE VALUES ... Core Value #1: Cypress is about winning. We will not tolerate losing. We thrive on competing against the world's best. Individuals can choose to compete and win in business at Cypress. Core Value #2: Cypress people are "Only the Best." We are smart, tough, and work hard. We tell the truth. We make no excuses. We value knowledge, logic, and reason. We deplore politicians. We are aggressive problem solvers who take ownership and get results quickly. Core Value #3: We "do what's right for Cypress." We are company owners. We choose "Cypress wins" over "looking good." We reward personal initiative. We are loyal and fair to our people. We follow the spec or change it. We keep commitments to customers. Core Value #4: We make our numbers. We gain market share, while we make excellent profit. We each set aggressive, quantitative goals--and we achieve them. We constantly improve. We do not tolerate waste. Core Value #5: We invent and build state-of-the-art products. Our technology challenges the world's-best head on, and we spend a fraction of what they do. Our first silicon always works on schedule. We manufacture at the world's lowest cost. We manufacture with excellent quality.

History of Cypress Semiconductor Corporation

Cypress Semiconductor Corporation designs, manufactures, and distributes a range of integrated circuits for the worldwide data communications, telecommunications, personal computer, and military systems markets. After starting up as a producer for the niche semiconductor products markets, Cypress turned its focus to mainstream chip markets in the 1990s. After its profitable SRAM memory chips became commodities in the mid- to late 1990s, Cypress turned its considerable research and development efforts to chips for mobile phones, wireless networks, and USB ports. Cypress prides itself on its entrepreneurial atmosphere. Its outspoken founder, president, and CEO, Thurman John (T.J.) Rodgers, is known as a tough boss of the "No Excuses Management" school--in fact, he co-wrote the book of the same name.

The General Patton of the Semiconductor Industry

Company founder T.J. Rodgers, who was named "America's Toughest Boss" by Fortune magazine in 1993, gained a reputation in the semiconductor industry for his visionary management, entrepreneurial and manufacturing techniques, and outspoken criticism of the U.S. semiconductor industry. Born and raised in Oshkosh, Wisconsin, his father a car salesman, his mother a schoolteacher, Rodgers displayed a talent for entrepreneurship at an early age. As a child, Rodgers learned electronics from his mother, who had taught basic electronics during World War II. In his high school chemistry lab, Rodgers created his first "product," a smoke bomb dubbed the "Thurm-O-Flare," a project which ended when Rodgers blew up the lab and put himself in the hospital. Rodgers was recruited to play football at Dartmouth University, where he earned bachelors degrees in physics and chemistry, graduating second in his class, and first in both majors. Accepted into Stanford University's physics program, Rodgers instead decided to study electrical engineering--despite having taken only a single semester of electrical engineering at Dartmouth--earning his masters and doctoral degree by 1975.

At Stanford, Rodgers studied under William B. Shockley, whose Shockley Electronics became the breeding ground of the famed Fairfield Eight (including Robert Noyce and Gordon Moore of Intel Corp.). While at Stanford, Rodgers invented a wafer etching technology called VMOS; he patented the technology, then sold it for cash and royalties to American Microsystems, Inc. (AMI), where he worked until he was fired--when the VMOS process was eclipsed by new technology--in 1979. In that year, the 31-year-old Rodgers attempted to start his own company--a goal he had set himself to achieve by age 35--but was unable to find financial backing. Instead, he accepted a position with Advanced Micro Devices (AMD), where he worked until leaving to found Cypress in November 1982.

Rodgers faced the daunting task of starting up--and arranging venture capital for--a new company during the recession of the early 1980s. Nonetheless, Rodgers attracted the interest of L.J. Sevin, founder of Mostek Inc., during the 1970s. After preparing a business plan for his proposed company, Rodgers, with Sevin's backing, began raising the $7.5 million in start-up funds envisaged in his business plan. Rodgers quickly found offers totalling $13.5 million, enabling him to choose his investors. As he told Santa Clara County Business: "We were therefore in the position of picking backers as opposed to begging for backers. Only the best! We picked backers that had 'value-added.'" Initial backers included Sevin, Sequoia's Don Valentine and Pierre Lamond, J.H. Whitney of New York, and Kleiner, Perkins, Caulfield, and Byers. In exchange for this backing of Cypress, Rodgers gave up 75 percent of control over the company. Cypress continued to attract venture capital, raising some $48 million in the years leading to its initial public offering in 1986. Cypress was underway by April 1983.

Cypress debuted its first product, a CMOS (complementary metal oxide silicon) memory chip, by the beginning of 1984. The chip contained the world's smallest and fastest CMOS transistors, which were 1.2 microns wide (compared to a human hair at approximately 100 microns wide) and operated three times faster than competing semiconductors while using 80 percent less energy. The company immediately set to work on a new round of CMOS transistors, shrinking them to 0.9 microns. By 1985, the company was profitable and posting revenues of $17 million. Following Rodgers original business plan--and hitting the targets envisaged therein--the company sought to operate solely within niche markets, those with sales of no more than $40 million. In this way, the company could avoid head-on competition with the semiconductor industry's U.S. and Japanese giants, which had neither the flexibility nor the interest in producing for such small markets. By 1987, the company was producing more than 70 types of chips, all targeted at high-end, sophisticated markets. Sales had already tripled to $55 million in 1986. Also in that year, the company's initial public offering raised $73 million, one of the largest public debuts for the time. By then, the company had added a second wafer fab, in Round Rock, Texas.

Cypress's extraordinary growth--by 1990 sales would reach $225 million, and net earnings of $33 million--was credited to an unusually intense corporate culture, beginning with its hiring process, a daunting marathon of some ten interviews. Rodgers quickly earned a reputation as the "General Patton" of the semiconductor industry, demanding that employees create weekly task lists, and putting into place a computerized system for tracking completion of the tasks employees set for themselves--and those who did not meet their targets could expect a withering dressing down from Rodgers. The company's manufacturing processes were equally as strict, enabling the company the flexibility to turn production quickly from one product to the next. As Rodgers explained to Business Month: "A semiconductor is a very unforgiving entity. If it takes 1,000 tasks to make one and you do 999 right but then forget one or do one wrong, the semiconductor won't work. Our system forces management to stick its nose in a big book every single week and find out what is going on. We can't afford surprises." In exchange for the high-pressure environment, Cypress employees received stock in the company. Rodgers, unlike many high-flying Silicon Valley CEO's, gave himself the same benefits--and imposed upon himself the same demands--as every other Cypress employee.

As Cypress grew, Rodgers also blossomed as an outspoken critic of the U.S. semiconductor, and particularly its quest, through Sematech and U.S. Memories, for U.S. government backing and intervention to help it compete against the rising Japanese power in the industry. Rodgers, a self-described Libertarian, instead demanded that the industry remain true to free market principles, a position that earned him little love among CEO's of his industry. Critics at the time pointed out that Cypress's focus on the niche markets--by 1990 the company was producing 159 products, generating as little as $1.6 million in sales each--enabled it to avoid the bruising head-to-head competition of the mainstream commodity markets. Nonetheless, Rodgers could hardly be faulted for the company's strong growth and its industry-leading productivity and profitability. At the end of 1990, the company bought its Bloomington, Minnesota, wafer fab from Control Data VTC, paying just $14.7 million. The company also opened the first two of its design centers, in Mississippi and Texas.

A Mid-Life Crisis in the 1990s

Static RAM chips (SRAM's) provided the bulk of Cypress's revenues. Still, the company began branching out into other territories, including PROM's and RISC processors, featured in, among others, Sparc processors for Sun workstations--a market that Rodgers predicted would eclipse the Intel-dominated personal computer market. As Cypress grew during the 1980s, Rodgers fought hard to maintain its entrepreneurial spirit. Rather than creating subsidiaries, Cypress instead began acting as a venture capitalist, spinning off divisions into separate start-up companies, wholly owned by Cypress, each with its own CEO and administrative and financial departments. Cypress's three subsidiary companies, which included Ross Technologies, created to produce high-performance RISC processors, contributed 10 percent of Cypress's sales by 1990; by 1991, these subsidiaries accounted for 28 percent of the company's revenues.

As sales continued to rise, reaching $286 million and record profits of $34 million in 1991, Rodgers strove to maintain the company's entrepreneurial origins. However, the company was headed for trouble: the SRAM markets had developed enough to attract the interest of the larger semiconductor companies, including Motorola, Micron Technology, Toshiba, Fujitsu, and Hitachi, sparking a price war that saw prices for SRAM's cut in half. With SRAM's moving into the mainstream, prices falling with the increasing supply, and a shrinking market facing a worldwide recession, Cypress stumbled into a mid-life crisis. Adding to its woes, Cypress had hitched its wagon to Sun's workstations. However, that market never developed into the Intel-killer Rodgers had hoped; worse, in 1992, Sun began reducing its inventories, cutting back on its orders to Cypress. Meanwhile, Cypress's Ross Technologies subsidiary was developing a new Sparc processor, with no guarantees that Sun would purchase it. Indeed, when Ross was late in delivering the new chip, Sun turned to Texas Instruments for supplies instead. Meanwhile, the personal computer market was moving into a new explosion in growth, and Cypress found itself on the outside of this crucial market. Rodgers's style of micromanagement was also criticized, particularly as the company was late with another important product, a one-megabyte SRAM chip, allowing his competitors to beat Cypress to the market. On top of all of this, Rodgers insisted on maintaining the company's assembling and testing facilities at its San Jose plant, despite the fact that keeping these activities in the United States cost the company some $17 million per year over what it would cost to move them overseas.

In 1992--to the delight of Rodgers's enemies, who had long chafed under his criticism--Cypress tripped up. Revenues shrunk to $272 million and the company posted its first loss in its history as a public company to the tune of $21 million. Chastened, Rodgers was forced to reorganize the company and adapt himself to Cypress's new conditions. Long pressured by his board of directors to move the testing and assembly facilities to Asia (although actual manufacturing would remain in the United States), Rodgers agreed, slashing some 700 jobs. It took Cypress only three weeks to move these operations. Rodgers also agreed to tone down his micromanagement of the company's activities in order to focus instead on its long-term growth. As part of the reorganization, Cypress brought its subsidiaries back into the company as divisions, while selling off the Ross subsidiary to Fujitsu for $23 million in 1993.

The glee with which Rodgers's detractors greeted the company's difficulties proved short lived. By 1993, the company was back on course, raising revenues to nearly $305 million and regaining profitability, with net earnings of $8 million. The following year, with an aggressive push into new product territories, including the personal computer market, Rodgers--who had taken the lessons learned over the last years to heart--once again led the company to new heights.

For the year ended 1994, Cypress posted sales of $406 million, and a net profit of $50.5 million. Aiding the companies return to growth were a series of acquisitions made between 1993 and 1994. The first, of IC Designs, Inc. for $16 million, brought that company's programmable clock oscillators for the personal computers; next, the company acquired Performance Semiconductor Corp.'s line of FCT-T high-speed logic chips, paying $5 million. In 1994, Cypress also acquired CONTAQ Microsystems Inc., and its line of personal computer chipsets, for $1.7 million.

Cypress arrived in the personal computer market to ride the surge in PC sales of the mid-1990s, spurred in particular by the introduction of the new Pentium processor, the first new PC chip since the introduction of the 486 at the beginning of the decade. The new computer market presented a double opportunity for Cypress--which began designing chipsets for the Pentiums--as the high-performance nature of the emerging technology opened a fresh demand for Cypress's core SRAM products. Where only 40 percent of personal computers contained SRAM in 1993, it was estimated that 80 percent would contain SRAM by 1997. Yet the company saw even greater benefits from the booming networking and telecommunications industries. By the end of 1995, revenues neared $600 million, and Cypress began making plans to become a $1 billion company, joining the industry's top ten, by the end of 1997.

In April 1996, the company began building a second wafer fab in Round Rock to meet the surge in demand. That project was delayed as the industry went through a worldwide slump in memory sales. Meanwhile, Rodgers's writing talents--he had also published a book, called No Excuses Management outlining his management philosophy in 1992--made the headlines in the summer of 1996. After receiving a form letter from a Catholic nun criticizing Cypress for not having any women or minorities on its board of directors, Rodgers wrote a six-page letter attacking such "political correctness" and the election-year hot potato of corporate responsibility and asserting that hiring on the basis of race or gender was immoral and "a lousy way to run a company." For this position, Rodgers was widely applauded among the executive and investor communities. Ever outspoken, in June 1997 Rodgers testified before a U.S. Senate subcommittee against "corporate welfare."

In October 1996, Cypress shut down its manufacturing facilities at its San Jose headquarters, converting the facility solely to research and development activities. Picking up the production was a second Minnesota wafer fab, and the start of construction on the new wafer fab in Round Rock, expected to be completed by 1998. At the end of 1996, however, Cypress found that its revenues had slipped back to $528 million in the face of the industry-wide slowdown. Though the company was planning to become a $1 billion company as early as 1997, it would not reach this goal until 2000.

Focusing on USB and Data Communications in the Late 1990s

Cypress restructured its manufacturing operations in the spring of 1998 to reduce its exposure to the volatile SRAM market. It stopped making memory chips at its plants in Round Rock, Texas and Bloomington, Minnesota, and also moved test operations from Thailand to the Philippines. SRAM production would continue at a second site called Fab 4 in Minnesota, where it was developing a 0.35-micron process. SRAM sales accounted for about 42 percent of revenues at the beginning of the year. Cypress was reluctant to exit the SRAM business altogether; it had over the years accounted for about 60 percent of the company's profits.

Sales slipped to $555 million in 1998, producing $111.2 million worth of red ink (after a $92.4 million restructuring charge). SRAM had become a commodity, and the Asian financial flu depressed sales in an important market. The semiconductor market, depressed for five years, began to show an upswing in late 1999. Nevertheless, turning its attention from the maturing PC market to data-communications applications proved the key to Cypress's recovery, noted Electronic Buyers' News.

Universal serial bus (USB) chips were another particularly exciting growth prospect. This technology made it much easier to connect personal computers with a variety of peripherals, from printers to joysticks. It also provided a much faster connection than parallel and serial ports.

After the Windows 98 operating system, which supported USB, was released, the market exploded beyond the expectations of all but the most astute industry observers. In May 1999, Cypress bought USB specialist Anchor Chips Inc. for $15 million and in September, it licensed advanced USB technology from rival Intel Corp., allowing it to enter the higher end of the market; Cypress already dominated the low end, with 20 million USB chips shipped. Cypress bought another USB chipmaker, Massachusetts-based ScanLogic Corp., in May 2001.

In March 2000, Cypress launched a new startup, Cypress MicroSystems Inc., to develop and market systems-on-a-chip (SOCs) for the eight-bit microcontroller market. SOCs were used in advanced data communications applications.

Cypress had sales of $1.29 billion for 2000, producing a net $277.3 million profit. The next year, sales slipped to $819.2 million and Cypress stomached a massive $407.4 million loss. Cypress made more than a dozen acquisitions in 2000 and 2001 as it focused on the fast-growing data communications business. By the end of 2001, Cypress's SRAM business accounted for less than half of total revenues, which were half of what they had been two years earlier. Nevertheless, the company laid off 14 percent of its workforce, or 650 employees, in mid-2001 due to a decline in orders. Most of the job cuts were in its Philippines fabrication operation, though 150 administrative jobs, mostly at the San Jose headquarters, were also shed.

As the company diversified, it reorganized into divisions based on market sectors rather than product offerings. In 2001, Cypress had four autonomous profit centers: memory, data communications, interface (USB), and timing technology. Four others were in the making: wide area networks (WANs), storage networks, wireless terminals (PDA's and mobile phones), and wireless infrastructures (base stations). By the end of 2001, Cypress had four divisions and 13 business units.

Entering 2002, Cypress's main focus in the datacom business was developing components for the latest generations of mobile phones. Next was supporting the Bluetooth wireless networking system.

Cypress created a new subsidiary, Silicon Magnetic Systems, in May 2002 to develop magnetic RAM (MRAM) chips using technology licensed from Minnesota start-up NVE Corp. Due to their small cell size, MRAM's promised to be powerful, inexpensive to build, fast, and durable.

Principal Subsidiaries: Cypress MicroSystems; Cypress Semiconductor (Minnesota) Inc.; Cypress Semiconductor (Texas) Inc.; Cypress Semiconductor Technology, Ltd. (Cayman Islands); Cypress Semiconductor Philippines Inc.; Silicon Light Machines Corporation; Silicon Magnetic Systems.

Principal Divisions: Memory Products; Data Communications; Timing Technology; Personal Communications.

Principal Operating Units: WAN (Wide Area Network); SAN (Storage Area Network); WIT (Wireless Terminals); WIN (Wireless Infrastructure).

Principal Competitors: Advanced Micro Devices Inc.; Integrated Device Technology, Inc.; Samsung Electronics Co., Ltd.; STMicroelectronics N.V.; Xilinx, Inc.

Chronology

• Key Dates:
• 1982: T.J. Rodgers forms Cypress Semiconductor.
• 1984: Debut of the company's first product--a CMOS chip.
• 1986: The company's IPO raises $73 million.
• 1988: Cypress lists on the New York Stock Exchange.
• 1990: A plant in Bloomington, Minnesota, is acquired.
• 1992: The company's first annual loss prompts restructuring.
• 1996: Memory chip production is moved from San Jose headquarters.
• 1999: Acquisitions and licensing deals bring Cypress into leadership of booming USB chip market.
• 2000: The company's sales pass $1 billion.