Current Issue
This Month's Print Issue

Follow Fast Company

We’ll come to you.

long read

Technology

The Untold Story Of The Invention Of The Game Cartridge

How a forgotten company's 1970s technical breakthrough launched a billion-dollar business and helped spawn a new creative medium.

Fairchild's Channel F, the first game console to use cartridges

Consider the humble video game cartridge. It's a small, durable plastic box that imparts the most immediate, user-friendly software experience ever created. Just plug it in, and you're playing a game in seconds.

If you’ve ever used one, you have two men to thank: Wallace Kirschner and Lawrence Haskel, who invented the game cartridge 40 years ago while working at an obscure company and rebounding from a business failure. Once the pair's programmable system had been streamlined and turned into a commercial product—the Channel F console—by a team at pioneering electronics company Fairchild, it changed the fundamental business model of home video games forever. By injecting flexibility into a new technology, it paved the way for massive industry growth and the birth of a new creative medium.

Almost two decades ago, cheaper means of distributing game software—first optical media, then the Internet—began to supplant cartridge technology. Even so, the business model created by Kirschner, Haskel, and engineers at Fairchild still remains as relevant as ever. Until now, their amazing story has never been told.

Fairchild's Channel F, the first game console with cartridges

The unwritten history of video game cartridges begins in an unlikely place: American Machine and Foundry (AMF), a manufacturer best known for its bowling alley equipment.

In the 1960s, AMF ran an R&D department in Stamford, Connecticut that originated many of its bowling technology innovations. Among them was an automated scoring system, devised in 1968, that kept track of individual player scores and displayed them on a video display mounted near the ceiling of a bowling alley.

Around 1969, AMF decided to move its R&D department from Stamford to North Carolina. Facing a difficult move, a handful of AMF R&D engineers led by a man named Norman Alpert decided to leave AMF and form their own company. Alpex Computer Corporation was born, named after Alpert himself.

At first, Alpex worked on an electronic cash register in partnership with mailing-equipment giant Pitney Bowes. But due to intense competition from IBM and NCR, the joint venture ended acrimoniously in late 1973.

Magnavox's Odyssey consolePhoto: Magnavox

Left dangling with no product to sell, Alpex Computer Corporation sought a new market for its talents. With a drastically pared-down work force of three employees (all refugees from an amusement supply company), the firm could afford to take risks. According to Wallace Kirschner, one of the ex-AMF engineers who followed Norm Alpert to Alpex, the company immediately considered entering the emerging video game market. Video games had exploded onto the scene less than two years before, driven largely by the success of Atari's Pong arcade machine and Magnavox's Odyssey home console (both launched in 1972).

Up to that point, all commercial video game products had been based on discrete, dedicated logic circuitry to generate on-screen gameplay. None in commercial production used a computer as the heart of its architecture. The Odyssey, developed by a team of engineers at defense contractor Sanders Associates, utilized mid-1960s technology that precluded the used of a computer. Likewise, the founders of Atari had originally sought to run games as computer software on minicomputers around 1969, but the high cost of the systems involved made that dream impractical.

The status quo changed in 1971, when Intel announced its 4004 microprocessor. On a single chip the size of a pinky fingernail, it contained the computing functionality of what would have once required multiple circuit boards full of electronics. The 4004 was followed by the 8008 in 1972 and the 8080 in 1974, each one adding dramatic leaps in capability over its predecessor. Competing semiconductor firms like Fairchild, National Semiconductor, and RCA began work on their own microprocessors, while engineers around the world applied Intel's new chips to just about every field imaginable.

It was in that climate that Wallace Kirschner realized that considerably more sophisticated video games could be generated using software running on a microcomputer with a bitmapped display rather than using dedicated hardware.

Lawrence Haskel in the 1970s

In early 1974, Kirschner got the go-ahead from Alpert to pursue his video game idea, but the firm realized it was racing the clock to develop this idea before competitors did. With that in mind, Alpex hired Lawrence Haskel, an engineer well versed in software who had worked alongside Kirschner at both AMF and Pitney Bowes-Alpex.

The plan was set: Kirschner would create the hardware, and Haskel, a big fan of games himself, would program the software. Their target? A home video game console which would let users switch games at will.

By 1974, the only home console the U.S. market had seen was the Magnavox Odyssey. It had been built primarily to play a game of video ping-pong, but could be expanded by using a series of jumper cards that turned on and off certain electronics inside the console. In tandem with taped-on-screen plastic overlays and physical board and card game paraphernalia, the Odyssey could provide new gameplay experiences. Even so, there are only so many truly fun games you can play by moving a couple of dots around on a TV screen.

Kirschner and Haskel began development on their game console in early 1974. The pair christened their new project "RAVEN," for "Remote Access Video Entertainment." At the heart of the system, the pair initially chose the Intel 8008, an 8-bit CPU. Around that, Kirschner built a device that could generate a 128-by-64 black-and-white display that used a pricey eight kilobits of RAM to store the image of the bitmap (the state of the screen's pixels).

Intel's 8008 chipPhoto: Wikipedia

"At the time, memory was very, very expensive," recalls Haskel. "I mean, a penny a bit, or something like that." That limited both the graphical capability of the system and the complexity of the software. Each game had to be less than two kilobits (or 256 bytes) in size. For comparison, this paragraph of text alone takes about 384 bytes to store electronically in its simplest form.

Reaching for inspiration, Haskel based his first program on the prevailing trend in the video game market: sporty, ping-pong type games popularized by the Odyssey and Atari's Pong arcade machine. The games made a big impact on Haskel, who vividly recalls the first time he saw the Odyssey in action during a visit to a department store.

"I was going to see the furniture department, and there was a little kid playing Odyssey," recalls Haskel. "I sat down and played with him for probably an hour. It was the coolest thing I'd ever seen. I couldn't get that out of my mind."

So for his first game program at Alpex, Haskel reinterpreted video ping-pong to reflect a cultural stalwart in the American northeast: hockey, which his family played every winter on a frozen pond behind his Connecticut house.

Haskel's electronic version of the popular sport looked Pong-like at first glance, with an overhead view of a hockey rink, two paddles, and a ball. Then he expanded on the Odyssey ping-pong model with on-screen scoring (a la Atari's Pong) and a game clock, and by allowing the player paddles (two white vertical lines on the screen) to rotate to deflect the ball in various directions. The main player paddle could also move forward and back across the rink. In a crowning touch, Haskel added a goalie paddle that could be independently controlled by each player.

"You had your hands really flying around with this," says Haskel of the game controls. The player could move a paddle up, down, left, or right; rotate it clockwise or counterclockwise; and move a goalie paddle up or down. There was a button for each movement.

"To this day, I still think that is the most challenging game that I've played," says Kirschner.

Haskel followed Hockey with Tic-Tac-Toe, a simple on-screen version of the game; Shooting Gallery, which allowed a player to rotate a "gun" and fire a bullet to hit a moving target; and Doodle, a primitive art program which let the player draw or erase white lines on the TV screen.

Needed: Switchable Software

As the game library expanded, it made sense to devise a way to interchange Haskel's programs, allowing players to easily switch between them inexpensively while using only one relatively expensive computerized console at the heart of the system. That way, the consumer would only buy the expensive part once and reuse it with a wide variety of software.

General purpose computers at the time approached interchangeable software with an array of removable storage options—often paper tape, magnetic tape, or packages of spinning magnetic disks—that required far too much expensive hardware to be practical in a consumer product. Kirschner and Haskel found a better solution right front of them, as a natural part of the development process. Intel's 8080 development kit encouraged the use of EPROM (Erasable-Programmable Read Only Memory) chips—a form of ROM chip that allowed the programmer to write and erase a program multiple times to speed up the development process.

Typically, once an EPROM was programmed, a hardware designer would either solder the chip directly to a printed circuit board or insert it into a delicate socket soldered onto such a board. It became obvious to Kirschner almost immediately that if consumers were going to use their console, they needed a way to change out those ROMs in a user-friendly fashion. So Alpex's engineers decided to mount the fragile ROM chip to a circuit board and, in turn, connect the chip's pins to a more durable connector that could withstand repeated insertion and removal.

That's how the first prototype video game cartridge was born.

"We went to RadioShack and bought these little plastic boxes," recalls Kirschner. "And we were able to plug the little box into the console with a connector we put on it." Kirschner remembers RAVEN's cartridge enclosures as being about five inches wide by three inches high by a couple of inches deep. Each black plastic box encased a circuit board with a memory chip containing video game code mounted on it.

These memory modules interfaced with the RAVEN console via a 25-pin connector protruding from the wide dimension of the cartridge box. Such a connector could withstand far more insertions than a delicate memory socket, but its 25 small pins were not suitable for consumer use. Alpex punted that problem down the road, setting the stage for a second round of video game cartridge innovation that took place at an entirely different company.

With all the key components of their RAVEN console in place, Kirschner and Haskel demonstrated their creation to Norm Alpert and Alpex's board of directors. Alpert knew that his small company did not have the money to market and manufacture the game console itself. It needed a business partner with deep pockets.

From its inception, Alpex had designed RAVEN to work with a common home television set as its display device. So Alpert first targeted American television set manufacturers as potential partners. In early 1975, his company approached Sylvania, Zenith, RCA, and Motorola (which once made TV sets), with Kirschner and Alpert giving demonstrations of the console under strict non-disclosure agreements. "None of them were interested," recalls Kirschner. "They had their heads in the sand, just coming out of the vacuum tube age, and they didn't do transistorized electronic designs themselves. That's why they later disappeared."

Enter Fairchild

Indeed, it took significant imagination to realize that Alpex's console, which existed then as a roughly 16-by-16-by-5 inch metal box filled with circuit boards, a handful of bulbous game modules, and an overly complex keyboard for input, could be streamlined into an attractive consumer product. Doing so would take major creativity, electronic design know-how, and marketing finesse.

With that in mind, Alpex shifted gears and approached semiconductor manufacturers, who were just then beginning to launch their own consumer electronics products. Among these gadgets were the first electronic clocks, the first pocket calculators, and early digital wristwatches.

In Alpex's early point-of-sale terminal days, a Fairchild parts rep named Shawn Fogarty had frequently supplied them with components. In early 1975, Fogarty received a very vague phone call from Alpert asking him to visit Alpex. Curious about what Alpert and his crew were up to, Fogarty agreed to visit the Alpex office for a demonstration. "These guys were smarter than shit," he remembers. "They were very capable and could put out real products."

Fogarty liked what he saw and sent news of it back Fairchild's vice president of Consumer Products, Greg Reyes, in California. The console made a positive impression on Reyes, who got the ball rolling with the CEO of Fairchild, Wilf Corrigan. Video games seemed like the next logical venue for the company's chips.

Corrigan tapped Gene Landrum, the former head of National Semiconductor’s consumer products arm, to be part of a team to evaluate the technology and write a feasibility report on the Alpex console from a marketing and business perspective. Landrum recalls the assignment: "Corrigan says, 'I want you to go to Connecticut, and I am going to send this engineer with you—this guy Lawson—and I want you to look at this thing and decide if we will buy it."

Fairchild engineer Jerry Lawson, in a photograph from the 1980s

That "Lawson" was Jerry Lawson (1940-2011), an electronics whiz from Queens, New York who happened to be one of the Valley's few African-American electronics engineers. By 1975, Lawson had already extensively experimented with video games himself, having created his own video arcade game in his garage around 1973. Knowing Lawson's affinity for video games, Reyes decided that Lawson would be an appropriate engineer to spearhead a new video game project within Fairchild.

"I had a secret assignment," Lawson told me during an interview in 2009. "Even the boss that I worked for wasn't to know what I was doing. I was directly reporting to a vice president at Fairchild with a budget."

The Alpex prototype used Intel's 8080 CPU—a chip from Fairchild's competitor (the pair had upgraded from the 8008 around the release of the 8080 in the spring of 1974). Ultimately, Kirschner and Haskel, with assistance from Lawson, had to convert their prototype into a version that worked with Fairchild’s F8 chip before the company would license the technology. Lawson also worked on taming RAVEN's awkward keyboard control scheme into a fairly complicated joystick that could work with Haskel's Hockey game. An industrial designer named Nicholas Talesfore created artist's concepts for the potential game console and hand controller, adding dramatic visual impact to Landrum's presentation.

Jerry Lawson's controller prototype

On November 26, 1975, Gene Landrum completed his report, "Business Opportunity Analysis: Alpex Video Game," for Greg Reyes. Accompanying the report was a schematic sketch of Lawson's controller prototype as well as Talesfore's initial designs for the production hand controller and game console. The project had picked up a new code name as well: "STRATOS". A portion of the write-up read: "This is an electronic video game aimed at the consumer home TV aftermarket. It is designed to eliminate the possibility of game obsolescence through the use of a 'unique' (and hopefully patentable) cartridge technique for adding additional games."

Among numerous statistics, Landrum projected that the STRATOS console could sell 5.5 million units by 1978, capturing 22% market share with a gross retail revenue of $220 million. They turned out to be highly optimistic numbers that never came to pass. However, they told Fairchild executives what they wanted to hear, and Corrigan gave the go-ahead for Fairchild to enter the video game business. The new console would be called the Fairchild Video Entertainment System, or VES for short (a name change would later re-christen it the "Channel F").

With Lawson in charge of electronic engineering and software for the project and Talesfore as head of industrial design, the group began work miniaturizing Alpex's unwieldy prototype into a size that would fit within a box that could sit comfortably atop a living room TV set. Before long, they realized that implementing the actual removable game software module would take special expertise. Talesfore knew just the guy to do the job: Ron Smith, a mechanical engineer he had worked with at National Semiconductor.

Only one year prior, while working at National Semiconductor, Smith mechanically engineered the world's first pocket calculator prototype with a removable memory module, the Novus 7100. But before the 7100 made it to market, National Semiconductor changed direction, eventually closing its consumer products division altogether.

As perhaps the only person on Earth who had previous experience designing consumer-oriented ROM memory modules, Smith was a qualified choice for the job, and Fairchild hired him as head of mechanical engineering on the project in early 1976.

"They had already started the tooling for the console," recalls Smith (referring to the process of making the metal mold that will later form the console's plastic shell), "and they were going to mold this box with no cartridge in it, then figure it out later."

Smith cancelled the tooling and hired additional engineers to make sure the fundamental internal mechanisms of the console were in place before the industrial design was finalized. "I said, 'Guys, you design from the inside out, not the outside in,'" recalls Smith.

After all, a key piece of the system was still missing: the cartridge itself.

Cartridge mechanism designer Ron Smith in the 1980s

Alpex's prototype had always included a way to exchange games via plug-in modules. But the modules were fragile and awkward. Fairchild had to envision a consumer-friendly way to package them, a job that fell primarily to industrial designer Nick Talesfore.

Inserting and removing socketed electronic assemblies had, until then, been an activity reserved for trained technicians, engineers, and military personnel. Taking a sensitive circuit board and putting it into the hands of a consumer—who might be prone to stepping on it, dunking it in the toilet, or leaving it baking in the sun—posed a considerable design challenge. Obviously, the board needed a protective shell of some kind.

8-Track Inspiration

Talesfore zeroed in immediately on the familiar form of the 8-track tape cartridge, an audio recording format which gained significant traction in the 1970s through its use in car audio systems. Relatively rugged, easy to insert and remove with one hand, and vibration-resistant, the 8-track tape proliferated where the comparatively delicate vinyl record feared to tread. He chose a shape and size for his new game cartridge enclosure that closely matched the 8-track tape standard. Then he added ribbing around the edges for improved grip, and selected a bright yellow plastic color to make a statement. Cartridges were the true star of the show, he figured, so they deserved to stand out.

Channel F cartridge (left) vs. 8-track cartridge (right)

Talesfore hired an artist he had previously met at National Semiconductor named Tom Kamifuji to design the labels. His illustrations depicted scenes that were often dramatically exaggerated (or confusingly abstract) compared to the actual on-screen graphics, setting a precedent that other companies would follow for years to come.

Each game package also came prominently numbered, starting with Videocart-1. This provided an easy way to refer to cartridges which contained multiple games of varying genres (Fairchild ultimately made it up to Videocart-26). In retrospect, the numbering system reflects a time when no one had any idea how many cartridges were enough for a system, how long a system like the Channel F could last on the market, or about the later appearance of third-party game publishers which could quickly balloon the number of games available into the hundreds, if not thousands. (Anyone up for a game of Videocart-963?)

As Talesfore worked on the graphic design for the system literature and the industrial design of the cartridges and the console itself, mechanical engineer Ron Smith tackled the real guts of the cartridge problem: how to physically attach and detach the cartridge from the console.

In an interview with Lawson I conducted in 2009, he touched upon the fears the Channel F team had in designing cartridge circuitry durable enough to withstand physical shock, thousands of insertions without breakage, and worst of all, electrostatic discharge. A stray zap of static electricity can easily fry an unprotected semiconductor chip.

"We were afraid—we didn't have statistics on multiple insertion and what it would do, and how we would do it, because it wasn't done," said Lawson. "I mean, think about it: Nobody had the capability of plugging in memory devices in mass quantity like in a consumer product. Nobody."

To protect from accidental damage, Ron Smith’s cartridge design contained a spring-loaded plastic door that protected a row of 22 gold-plated contacts inside the plastic housing while not in use. Once inserted into the console, a cam would open the door, exposing the contacts. Then a special connector with flexible metal pins inside the console would rotate up to meet the cartridge and make electrical contact—while simultaneously sweeping them to break through any corrosion that had built up on the fingers.

As he worked, Smith took notes—including these pages dealing with the chute which the console used to hold the cartridge while in use:

  1. 01/04
  2. 02/04
  3. 03/04
  4. 04/04

Finally, a separate mechanism would lock the cartridge into place so that it would not jostle loose while active; it could be released with an eject button located in front of the cartridge chute. This frictionless, zero-force experience ensured that the consumer would never struggle with inserting or removing a cartridge from the system. As with an 8-track tape, you could do so using one hand if necessary.

Smith and Talesfore later received a patent for their work on the Channel F console and cartridge, providing a basic template for cartridge-and-console design that many game systems would follow for years to come.

Here are drawings from some of the patents associated with the Channel F project:

  1. 01/05
    Schematic for Alpex's RAVEN proto-console
  2. 02/05
    The Channel F system, complete with TV
  3. 03/05
    The Channel F game controller
  4. 04/05
    The cartridge
  5. 05/05
    The cartridge-chute mechanism

The Debut

The Channel F made its first public appearance in June of 1976 at the Summer Consumer Electronics Show in Chicago. However, the company only displayed a non-working empty shell, so the system did not attract much press attention. A few weeks later, the Channel F made a much bigger debut nationwide as part of a July 6, 1976 article in Businessweek called "The Smart Machine Revolution." The multi-page feature touted the Channel F alongside cars, watches, and scales as a product that demonstrated the enormous potential of microprocessors in everyday consumer products.

Following the article, excitement for the concept behind Fairchild's new system rippled through the home video game industry, which faced a glut of dedicated systems by that time—almost all of which played a copy of Atari's Pong. The new razor/razor blade business model embodied in the console/cartridge system represented a welcome change, with video game revenue forecasts from then on projecting massive growth of the industry due to sales of high-profit cartridge-based software that played on low-profit consoles.

Most importantly, the debut of the Channel F served as a wake-up call to those companies who, by 1976, also had cartridge-based consoles in the works. RCA was one of them with its Studio II, which was developed concurrently with the Channel F. William Bachman, one of the co-inventors of the Studio II's cartridge mechanism, confirmed in an interview that the system was in the works at least by May of 1976—before Fairchild went public with its console.

Atari, which had also been working on a programmable console concept, received physical cartridge design inspiration from a familiar place. In 1976, Ron Smith had hired a contract engineer named Doug Hardy to assist him with designing the Channel F cartridge mechanism. Before completing the project, Hardy quit the job and joined up with Atari. One year later, Hardy co-designed the cartridge shell and mechanism for the Atari VCS (later known as the Atari 2600), which would debut in October 1977. While the 2600 cartridge bore a different internal construction to avoid patent trouble, the basic cartridge metaphor pioneered by Fairchild had already made its mark.

A brochure for the Channel F

Shortly before launch of its console, Fairchild changed the name of the console from the Video Entertainment System (VES) to the Channel F—short for "Channel Fun." Unfortunately for Fairchild, the slight first-to-market lead for the Channel F did not confer any particular advantage to the console in the long run. With titles like Math Quiz, Video Blackjack, Mind Reader, Maze, and Acey-Deucey, Fairchild's games tended to be light on action. "We learned at Atari that if you weren't pounding someone and shooting someone and blowing someone up, people didn't like your game that much," says John Ellis, former VP of consumer engineering at Atari. "Not to say that I agree with that philosophy, but that's the way it was."

RCA also learned the hard way that taking the intellectual path was often bad for business. Its almost comically primitive Studio II system (which launched a mere two months after the Channel F in February 1977) relied heavily on simplistic educational titles and slow sports like bowling (arguably its best game). It tanked within a year of its launch in great part due to the fact that its titles were arguably more boring than watching paint dry—although with black and white graphics at 64-by-32 pixel resolution, there wasn't much room for anything more exciting on the screen.

The Channel F sold fairly well compared to its dedicated console predecessors during its first year on the market, but the Atari 2600, with its library of vibrant action titles and home versions of popular arcade games, soon overtook it, then completely engulfed it.

Still hanging on, in 1978, Fairchild redesigned the Channel F to give the console a fighting chance against the 2600. The Channel F System II came with a sleek new look and new detachable hand controllers which included bona fide action buttons. But it was too little too late for Fairchild, and the revised model only saw limited release before the firm decided to pull the plug on the system altogether, selling the technology to Zircon International in 1979.

Under Fairchild's stewardship, the Channel F sold only around 350,000 units total. During that same period, Atari moved millions of units of the 2600 and millions more software cartridges.

Here are images of some of the Channel F's games, taken from an advertising brochure:

  1. 01/07
    Baseball
  2. 02/07
    Drag Race
  3. 03/07
    Maze
  4. 04/07
    Blackjack
  5. 05/07
    Space War
  6. 06/07
    Math Quiz
  7. 07/07
    Spitfire

Game Company Beats Chip Companies

Why did Atari beat Fairchild and RCA? It specialized in games. It wanted to entertain, and it wanted to do so vividly, bombastically, viscerally. Its experience developing midway-style arcade titles had given it a taste for what attracted and what pleased, and its laid-back management style had fostered innovation at an elemental level that fed its explosive growth through the 1970s.

Contrast that with Fairchild and RCA, which only operated their consumer divisions as vehicles for their semiconductor chip businesses. For upper management, fun was almost beside the point.

To that end, the Channel F suffered badly from internal corporate maneuvering. 100% of the chips in the Channel F came from Fairchild Semiconductor, which forced its sister division Exetron (the consumer division responsible for the Channel F) to buy the chips at a fixed price (which was sometimes higher than the price it offered to competitors), gaining a profit from each sale.

Meanwhile, Atari could pick and choose the lowest market prices for its chips from a wide variety of vendors. With no semiconductor overlord to please, Atari was fast on its feet, allowing the company to reduce the cost of its console as much as possible to undercut the competition.

Ultimately, the games-as-software model allowed the possibility of selling a base console on a slim margin (or even a loss) to achieve as high a market penetration as possible, then making up the loss from sizable profits of relatively inexpensive-to-duplicate software. This same business model is what drives the video game industry today, almost 40 years later.

Today, the creators of the first cartridge-based video game system have little to do with the video game industry. But their contributions ushered in a new era. Making games interchangeable allowed consoles to transcend their hardwired origins, giving birth to the video game as a creative medium with endless potential. In that sense, the cartridge was as much as cultural revolution as a technological one.