How I Started

Disclaimer: Opinions expressed herein are my own and are not intended to represent those of organizations other than Syncopated Systems.

My Early Years

My parents met while both working for IBM in the Netherlands (where my mother is from); they encouraged my interest in science and computers, which started in the 1970s when I was still in elementary school, probably when I visited my father’s office and he showed me how to type my name on a card punch. He also loaned me a book on COBOL (which I read only a few of chapters of at the time) and introduced me to electronics through his collection of American Flyer toy trains, Radio Shack kits, local surplus stores, and circuit assemblies discarded by his employer, Control Data Corporation (CDC), where Seymour Cray created the first commercial supercomputer.

The First Digital Computers

One of my father’s colleagues, LaFarr Stuart, in early 1962 had gained a national radio audience as a graduate student at Iowa State University for having programmed its early digital computer (Cyclone, a derivative of ILLIAC) to play simple, recognizable melodies. He started teaching me about digital logic using his first-edition Texas Instruments TTL Data Book (an unusual hardcover data book) and a Digital Equipment Corporation logic trainer kit (probably intended for college students), and about assembly languages and machine code using a COSMAC Elf single-board computer (with a whopping 256 bytes of RAM).

In the late 1970s, computers with typewriter-style keyboards were new and I was eager to use them. Classmates introduced me to computers in their homes including a Data General Nova (a minicomputer in an equipment rack under one side of a desk) and a Durango Systems F-85 (a desktop microcomputer nearly the size of a desktop). Around 1980, my elementary school tested my intelligence (IQ) and put me in a program for “gifted” students, which had a single shared Apple II computer.

Atari

In the next few years, I learned to program extensively in BASIC mostly on Atari 800 home computers. I started repairing and upgrading them; as Tandy Corporation acquired the home electronics retail chain VideoConcepts in 1985, I bought its remaining Atari-related inventory—including its field service kit—and started advertising online. I started attending meetings of computer users groups, including the Bay Area Atari Users Group (BAAUG) and one attended by a Stanford University math professor who introduced me to the IBM RT PC workstation in his office, probably in 1986. In 1987, I worked a summer job at Atari, where I had conceived and suggested a system to create a local area network (LAN) with Atari ST computers.

In February 1988 and only about six months after I had worked at Atari, across the street from it was a horrible mass shooting at ESL. Two friends of a friend died in the shooting.

About five years later, the 1993 made-for-television film I Can Make You Love Me starring Richard Thomas and Brooke Shields was made about it and broadcast while I was enrolled in a class taught by one of the officers who had been called to the scene of the rampage.

Serial Data Communication

Starting in 1985, while using a modem I became very familiar with the methods they used for serial data communication (mainly the EIA/TIA-232 standard, commonly known as RS-232). Tandem Computers, where my father worked at the time, at one point offered surplus equipment to its employees, so I got a couple of data terminals (and other equipment) that I experimented with.

(During my last year of high school, I also won an award for a Pascal program I had written on an Apple IIe that emulated a computer bulletin board system I accessed with that modem.)

In college, I found that I could conveniently emulate an IBM 3278 data terminal to communicate with an IBM 3081 mainframe.

Shortly after Maxim Integrated Products released its MAX232 integrated circuit in 1988, I used them and began learning about voltage-level translation and switched-mode power supplies—working much more closely with both of these only about five years later. I had started programming with serial data streams in the 1980s; as GPS became commercially available in the early 2000s, I wrote software that read serial data from GPS receivers, and have used my familiarity with GPS on projects since then.

High-Performance Computing

As I entered college (at first majoring in engineering), I worked at IBM’s Almaden Research Center, where I had worked with the IBM RT PC and learned a lot, especially while helping to develop a rack-mountable chassis for its internal use. In college (which I continued to work through most of), I learned to program mainframes (IBM and CDC), Unix-based minicomputers (Sequent and IBM), and use the languages COBOL, Fortran, C, and C++.

Another of my father’s CDC colleagues, Dwight Chang, for several years with my father had run a side business (unrelated to technology) and encouraged my interest in high-performance computing (HPC) and in IBM RS/6000 computers.

In the early 1990s, I became a licensed developer for the Atari Portfolio palmtop computer and IBM RS/6000 computers, which succeeded the IBM RT PC. I also became old enough to buy beer.

My father, Rick Carlsen, died in 2018. Two of his colleagues from Control Data Corporation, Dwight Chang and LaFarr Stuart, became my friends and mentors, and both died (coincidentally) in 2021, at ages 75 and 87, respectively.

A Tale of Two Startups

In 1978, my family moved from San Jose to Sunnyvale—putting us between and within only about four miles (6.4 km) of both Apple Computer (as it was named until 2007) and Atari.

Today, clusters of local buildings bear Apple’s signs. (Apple also occupies other buildings discreetly.) I recall seeing the construction of:

I also recall seeing a similar cluster of Atari’s buildings in north Sunnyvale, and—after the video game crash of 1983—in 1984 seeing Atari split into two companies. The consumer products company, Atari Corporation, was reduced to three buildings, which included two warehouses and its long-time headquarters (at 1196 Borregas Avenue), where I had worked a summer job in 1987.

Now nearly 35 years later, Apple has become the first United States public company to be valued at one-, two-, and three-trillion dollars. After I worked for Atari, it continued to operate for only 9 years more; shortly before this writing, Atari’s headquarters building had been reduced to rubble.

The Co-Founders

In 1971, Nolan Bushnell and Ted Dabney adapted the 1962 computer game Spacewar! to create the first commercial video game (and first arcade video game) Computer Space, which was not commercially successful. In 1972, they co-founded Atari by essentially stealing an idea from the Magnavox Odyssey and hiring engineer Al Alcorn to make Pong. Bushnell bought out Dabney in 1973 (for $250,000), settled a lawsuit by Magnavox in June 1976 (for $12-million), sold Atari to Warner Communications in November 1976 for an infusion of cash that would allow it to compete with the first game console with replaceable program cartridges (Fairchild Channel F, released in 1976), and continued to run Atari until 1978.

Around when I started, a best friend and I had been restoring a coin-operated Pong game he got from our high school’s electronics lab, but my father wanted it out of his garage. After learning from my colleagues that Bushnell had given his last unit to the Smithsonian national museum when it asked for one, I gave him my Pong game for Christmas, and he particularly enjoyed the familiar names he found in the original documentation, which had been in the bottom of the cabinet. (Through this, I also met Alcorn when he happened to visit the office, and have since run into him a few more times at local IEEE functions.)

At first, I helped develop a system of customer-facing point-of-sale terminals with large color touchscreen displays (some of the first) and autonomous mobile robots that ultimately (and for a relatively short time) delivered food and drinks to customers’ tables at a Little Caesars pizza restaurant near its corporate headquarters in the Detroit area. At one point, for our client we created a demonstration video at the restaurant and bar The Lion and Compass, which was created by Bushnell’s wife Nancy Bushnell and her brother Robert Nino in 1982 and where I dined on its closing night in 2017. We had subcontracted the creation of the imagery for the terminals to Pacific Data Images (PDI), which I toured on a field trip through a college course in computer animation before its office moved from Sunnyvale in 1995 and before it began creating feature-length films and became part of DreamWorks SKG in 2000.

I bought my first modem, a used MPP 1000C, from my friend Dave Miller. He was about eight years older and had worked for Bushnell (at Axlon), years before I did. Dave Miller died in 2012, leaving behind his wife and two teenage sons.

I recall Bushnell taking us out to lunch at a Chinese (dim sum) restaurant that occupied the building where he started the first Chuck E. Cheese location when I was still in elementary school. Circa 2001, the building was razed to construct the Santana Row mixed-use development.

Bushnell had the first in-car navigation I had ever seen in operation, made by a company that he had funded called Etak. (I also knew the founder’s stepson, who also had one in a neat old Volvo that apparently had been a hand-me-down.) Etak’s cars had driven the streets creating digital copies of street maps; as the world-wide web gained popularity, Yahoo! leveraged Etak’s map data.

When the Loma Prieta earthquake hit in October, I was contributing to the development of the first television adapter for personal computers, Aapps Corporation’s MicroTV (for Apple’s Macintosh II). By the fall of the Berlin Wall in November, I had followed the company as it had spun out to its own office, where we watched the event using a large portable (C-band) satellite dish.

Not long after I stopped working for Bushnell, he had laid off his long-time chief operating officer, Chuck Onorato, who died shortly afterward in 1990. He had a larger-than-life personality, and I had often thought that he might have been inspired the name of Chuck E. Cheese. His funeral is the first I recall attending as an adult.

I had also worked with his nephew Greg Onorato, who was closer to my age though still a few years older and had worked with Dave Miller at Axlon. We reunited in 2012 when stepson Erik Onorato died; despite his mourning, his personality was just as warm as I had remembered him. Greg Onorato died in 2019.

Through Atari, Bushnell had briefly employed Steve Jobs, who in 1976 co-founded Apple. In 1975, Jobs had offered to sell Bushnell one-third of Apple’s ownership for an investment of $50,000 (equivalent to roughly $316,000 today). Had he taken the offer, Bushnell likely would have become the world’s first trillionaire and the seventh-wealthiest person ever.

Innovations and Imitations

Prior to Apple’s initial public offering in 1980, Jobs had traded some of Apple’s stock to Xerox for technology it had created in Silicon Valley at Xerox PARC (Palo Alto Research Center) since its opening in 1970; I once estimated that this stock would have grown in value to about one-billion dollars—roughly one-twelfth of Xerox’s net worth—when Xerox was acquired by Fujifilm in 2018.

Apple used this technology to create its Lisa and Macintosh computers (released in 1983 and 1984 respectively), and their revolutionary graphical user interface, laser printers, and data networks.

Though Microsoft hired at least one notable researcher from Xerox PARC (Charles Simonyi, who I met in 2019), Microsoft merely imitated earlier graphical interfaces (the X Window System and its predecessors including Macintosh) to release Microsoft Windows in 1985, as it had imitated CP/M with MS-DOS in 1981 and imitated Digital Equipment Corporation’s BASIC-PLUS to create Microsoft BASIC in 1975.

Exits

In 1989, Businessland had negotiated to sell NeXT computers and sponsored a kick-off event at which Jobs would speak. Bushnell’s office received an invitation, which no one else wanted it, so I attended. After his presentation, I had approached Jobs to thank him for the invitation, but with the press all over him, after a few minutes moved on.

Atari Spawns an Industry

During the first three years of Kassar’s leadership, Atari’s annual sales increased nearly 30 times: from $75-million in 1977 to over $2.2-billion.

Despite his early gains, Kassar disregarded the value of the company’s game programmers and referred to them (at least as early as 1979) as “high-strung prima donnas“ who were “no more important to that game than the guy on the assembly line who puts it together.” (Imagine if someone running Warner’s music division had said something like that to a recording artist, or if someone in its motion picture division had said similar to a director or someone playing a leading role.) Apparently as a result, during the same three-year period, nearly all of Atari’s original staff were fired or had quit, some starting the first two third-party game developer/publishers Activision in 1979 and Imagic in 1981 (the latter operating only until 1986).

I worked for Activision in 1990 to 1991, when it was called Mediagenic. (An inside joke there was that the name meant “looks good on paper.” Though it might, the implied joke was that it looked good only on paper—as planned, but not as implemented.)

I knew someone who worked in the test lab, and that’s where I started—at first romanticizing the idea of getting paid to play video games all day and in the first week already having nightmares about being stuck playing through the same game endlessly. After a few months, I managed to move up to more interesting work developing Sargon V: World Class Chess with the creators of the original Sargon chess program including Kathe Spracklen, who was a joy to work with.

In addition to contributing to game development, I had joked about the low price of the company’s stock by expressing the cost to buy a share as the number of soda cans one could return for their recycling deposit; I recall it reaching a low of $0.10 (at the time, four cans) per share and a couple of occasions when someone bought large blocks of stock and denying on the required forms any intention of acquiring the company. I accepted another job and gave my customary two-week notice only about a week before the afternoon on Friday February 1, 1991, when an unscheduled all-employee meeting was called; when we had all been corralled into the multi-purpose room, CEO Bruce Davis announced that the company have been acquired and introduced the new owners as they walked in, including Bobby Kotick.

Two weeks later (and a week after I changed jobs), Activision laid off about 90% of its workforce, reducing its size to about 30 employees; about 10 of those remaining quickly left on their own. Though this was unfortunate for my colleagues, many quickly found work at other companies including nearby Sega of America. (Nearly three years later, Sony Computer Entertainment America was formed nearby to create PlayStation, which was released in December 1994.)

Going, Going, Gone

After I had gained experience working at a few startup companies and then returned to college to study business, I was somewhat surprised to learn of a widely-accepted accounting principle (though not formally one of the Generally Accepted Accounting Principles in the United States) called the “going concern assumption,” which seems to imply that a business should follow a trajectory of success that is roughly horizontal at some arbitrary altitude, at least for the limited future that near-sighted accountants concern themselves with.

In contrast, Atari’s history demonstrates the limitations of the going concern assumption through its example of the lifespan of a Silicon Valley business; it has clearly-defined periods, beginning with steady growth from its formation in 1972 to 1978 while led by Bushnell, rapid growth from 1978 to a spectacular crash in 1983 while owned by Warner and led by Kassar, and ends with a steady decline from 1984 to its demise in 1996 while led by Commodore International founder Jack Tramiel.

Though I had preferred Atari to Commodore, I learned to admire Jack Tramiel for what he had built. While I worked for Atari, I had also seen the number that had been tattooed on his forearm when he was in a concentration camp, and younger than my age then. Jack Tramiel died in 2012.

Long separated from those who built them, Atari and Activision exist today as little more than “zombie brands.”

Digital Audio

Now that smartphones are common, it’s easy to forget that they have been around for only 15 years, since Apple introduced its iPhone in 2007.

Before that, the most popular portable computing devices were MP3-based digital audio players, especially Apple’s iPod, which was released in 2001.

SigmaTel

Though I hadn’t contributed directly to the development of the iPod, while working for Austin-based chip maker SigmaTel in 2000-2001, I did create a new user interface and a reference design that were used to create most of the flash-based MP3 players made in 2004 plus Apple’s first-generation iPod Shuffle, which was released in 2005 (using SigmaTel’s second-generation MP3 player system-on-chip) and reached a peak production rate of 100,000 units per day.

It was only about 10 years earlier that most personal computers became able to play high-quality audio.

Media Vision

In 1991, I joined Media Vision, a company with about 15 other people and a first product that was nearly complete: a PC sound card called Pro AudioSpectrum that included a digitally-controlled analog mixer and a SCSI interface for a CD-ROM drive, which were still relatively new and were available packaged in kits with some of the cards. (The kit with the internal Sony CDU-541 drive cost $800 and the kit with the external CD-ROM drive cost $1000.) Next came the Pro AudioSpectrum Plus, which added full compatibility with the Sound Blaster sound card, followed by the Pro AudioSpectrum 16, which increased sampling and playback resolution from 8 bits to 16 bits, enabling the same high-quality audio as compact discs. For the latter, we produced 100,000 units in its first month; years later, when I mentioned this to an interviewer in Apple’s iPod group (while inside Apple’s first building, which I saw constructed as a child), they laughed at the small number, obviously not realizing how much more complex the products had been to build than a modern iPod.

I had joined Media Vision only about two weeks before its first major trade show; because the company was small and I had interviewed while wearing a suit (revealing that I had one), the company put me in its booth. Down our row was Nolan Bushnell in the Commodore International booth promoting its new CDTV media appliance.

With Media Vision’s first product, my manager introduced me to the process of getting FCC certification as a Class B computing device regulated under 47 CFR Part 15, and made it my responsibility for subsequent products. Though this kind of interference testing could be done in urban areas using a shielded chamber (such as one I had seen years earlier in one of Atari’s labs), we outsourced most of the process to a company that specialized in testing and certification located in rural Morgan Hill (about 30 minutes south of San Jose) that tested equipment on rotating tables in a grassy field. I knew our products were going to be a hit when one played a piece of stock digital audio of a rancher herding cows and the cows in the next field came closer, raised their heads, and started mooing too.

Though I had many outstanding colleagues at Media Vision, I’m particularly proud to have worked with John Neary (1964-2015). He was the lead hardware engineer of the AudioPort product (for playing and recording audio with laptop computers through a parallel printer port), for which I recall introducing him to the FCC testing site (as my manager had earlier introduced me). He and his wife on their honeymoon had met at their hotel the great comic actor Leslie Nielsen; I found much of the same humor in John Neary. Later, he went on to Dolby Laboratories and received (with his colleagues there) a 1997 Oscar award and a 2009 Emmy award.

Another notable thing I did at Media Vision was supervising the first pressing of Microsoft Windows onto CD-ROM in 1991. After bringing the first 500 copies back from the factory (plus 500 copies of the pack-in game Jones in the Fast Lane), one of our inside sales people mentioned taking an order for two copies from Bill Gates’s assistant.

Though I worked in the engineering department, when production problems occurred, the executives called me in to resolve them, and I did. At one point, my manager (the director of hardware engineering, who had been an early Apple employee and a good technical mentor to me) mentioned his plan to make me exempt from overtime pay (without giving me a title promotion) at a rate that would have substantially lowered my pay rather than rewarding me by increasing it; so, I quit and sold my stock in the company as soon as I was able, coincidentally as its price peaked. Soon afterward, Media Vision apparently was no longer able to meet its aggressive development and production goals, started cheating on its books by declaring unfinished goods as shipped (and hiding them off site), and became the subject of the longest-running securities fraud case in Silicon Valley history.

More Video Games

After I had worked for both the company that built the early video game industry and its co-founder—Atari and Nolan Bushnell—followed by the first independent video game developer and a diversion into digital audio through Media Vision (which also became a game publisher), I got back into game development.

Iguana Entertainment

Shortly before I started to work for Media Vision, I started working evenings and weekends (“moonlighting”) helping a friend at a small video game developer Punk Development (which I later learned was part of game publisher Razorsoft). It occupied a suite with a lobby, three offices, a restroom, and a small rear warehouse in a Sunnyvale industrial park across the street from The Lion and Compass. When it dissolved, many of its former staff established video game developer Iguana Entertainment and I was considered one of its founders (listed in its articles of incorporation, as I learned later); in those offices, I saw my incredibly-talented colleague Matt Stubbington paint the original logo on canvas.

Matt Stubbington died in 2011 at only age 39, and to me embodied Lao Tzu’s “The flame that burns twice as bright burns half as long.” We shared love for music and similar tastes, and together saw great acts including Seal in the mid-1990s and Duran Duran in 2007.

Though Matt Stubbington’s death may have hit me the hardest, I was also deeply saddened by the deaths of our friends and colleagues Carl Wade in 2012, J. Moon in 2018, and Gregg Hargrove later in 2018

In early 1993, I participated in a project called Joint Venture: Silicon Valley during its initial set of eight meetings: one for each segment of local industry—mine being for semiconductors. As a “student observer,” I sat behind the mayor of Santa Clara. To his left a few seats was AMD co-founder and CEO Jerry Sanders, who had already moved much of his operation to Austin, Texas; much of the discussion that evening revolved around why companies were moving there.

Later that year, Iguana Entertainment was looking to lower its operating costs by moving out of California. I shared what I had learned and moved with it to Austin, becoming Iguana Entertainment’s Senior Hardware Engineer.

As a one-person department reporting to the company’s president, I continued to reverse-engineer video game consoles to create devices that enabled efficient development of game software by providing interfaces to personal computers (originally Commodore Amiga computers followed by IBM-compatible PCs) and emulating game distribution media, at first ROM cartridges and later CD-ROM. (While at Mediagenic, I had seen it develop games using commercial general-purpose ROM emulators—which were too expensive for us— and some of the original custom-made development hardware used to create Activision’s groundbreaking first games for the Atari Video Computer System, which I later saw for sale at a local surplus shop.)

I also created software tools, including the company’s first programs for Microsoft Windows, which were used to create NFL Quarterback Club 96 and its subsequent annual versions.

One of the more interesting things I created at Iguana Entertainment was the latest version of my development system hardware for the Super Nintendo Entertainment System, in which I emulated ROM inside a battery-powered cartridge that could be used independently by game testers. These I created by using a relatively new type of memory called pseudo-static RAM, which had significantly lower cost than the static RAM (SRAM) that I had used in prior designs, but required a power supply that was much more regulated. For this, I created a very small switching power supply regulator using new controller chips as they became available from Maxim. (Good inductors were not yet widely available through distribution, so I had to buy direct from a manufacturer, and many more than I would ever use. Fortunately, they were still relatively inexpensive.)

Sony PlayStation

Another of my favorite projects at Iguana Entertainment was reverse-engineering Sony’s PlayStation to create a development interface. I had removed components from the circuit board including the system ROM, for which I wrote software to extract and disassemble. (Though the first version of my interface needed to boot from a CD-ROM, my second version booted even faster from an on-board EPROM.)

In 2006, I helped Sony’s American PlayStation division (then still named Sony Computer Entertainment America) in testing and demonstrating a new API for the Blu-ray disc drive on the PlayStation 3.

In 2010, Sony wanted to develop a PlayStation for a niche market. After what I had done at Iguana Entertainment, I apparently knew the original system better than anyone outside of Japan. So, Sony hired me as a Senior Staff Hardware Engineer and I developed working prototypes for a product projected to produce $1-billion in retail sales. (Peak production would be near 20,000 units per day.) I demonstrated these at Sony’s headquarters in Tokyo during its annual company-wide technology exchange event to (among many others) Kaz Hirai, who was then my fourth-level manager and later became Sony’s CEO and Chairman.

Some of my more interesting colleagues at Sony included R.J. Mical (formerly of Amiga) and Jerry Jessop (formerly of Atari).

Although Sony has fewer employees now, Sony was (and likely still is) the world’s fifth-largest media conglomerate. Sony still has more than 100,000 employees and at the time had roughly as many as Google.

Another Rack-Mounted Computer

After leaving SigmaTel in 2001, I leveraged what I had learned while creating switching power regulators at Iguana Entertainment at another growing Austin startup called Newisys. There, I budgeted, designed, and sourced major vendors for the power subsystem in the first computer with the x86-64 microprocessor architecture (introduced by the AMD K8 Opteron), which—even more than 20 years later—is still used in nearly all desktop, laptop, and server microcomputers.

At Newisys, my second-line manager was Phil Hester, who had earlier led the development team of the IBM RS/6000 series of computers (for which I had become a licensed developer many years earlier). He had later become CTO of AMD before his death in 2013 at age 58.

Rail Equipment

More than two years after I left Newisys, I developed a product for a nearby rail equipment company. At first, it had contracted with an engineer who had later joined Newisys and could not get the control interface working reliably, so I was tasked with correcting his mistake. (Although I dislike criticizing the work of others, his mistake was inexcusable.)

I soon also made the interface about ten times smaller and lighter, and ultimately created a system that was remarkably efficient and profitable, dominating its niche market.

Among other things, I had given the system real-time artificial intelligence, which is common in video games.