https://tedspence.com/the-rise-and-fall-of-6502-gaming-31b84e2d0cd?gi=566d1df6c491 Open in app Home Notifications Lists Stories --------------------------------------------------------------------- Write tedspence.com Published in tedspence.com Ted Spence Ted Spence Follow Jun 29, 2020 * 13 min read Save The rise and fall of 6502 gaming The life of the computer chip that sparked a revolution in gaming About forty five years ago, a team of developers at Motorola had a great idea for a new microprocessor: one that would be simple, easy to use, and less expensive to build. Motorola wasn't interested, but they built something amazing anyway, and the legacy of this small team lives on in every personal computer. The low cost of the 6502 also had unexpected side effects. Within three years, it created a completely new market: home console gaming. How did this tiny chip rule our living rooms for over a decade? We're all making it up as we go along It was early 1975. Itching to build a great new product, a band of engineers decided to quit Motorola and design a new CPU. The most talented salesman in the bunch, Chuck Peddle, talked his way into a deal: they would join John Pavinen's MOS Technology, a small firm all the way out east in Pennsylvania. Microprocessors were still new; they had been introduced commercially in 1971 by Intel. The entire commercial microprocessor market was less than four years old and mostly led by Intel, Motorola, and Texas Instruments. Whereas Motorola and Intel were earning hundreds of millions of dollars per year, MOS Technology was so small it would be sold to Commodore barely a year later, for a fire sale price of $12 million. Yet at small companies, the struggle to survive often leads to more creativity. MOS engineers had developed a technique to fix up damaged chip masks called spot knocking. During chip fabrication, this technique saved huge amounts of money and increased chip yields dramtically. With the fate of the company at stake, these scrappy chip designers could iterate on a design faster than anyone else. The entire 6502 design was completed, under the leadership of Bill Mensch, in less than six months. [0] The 6502 chip (IEEE Spectrum) But what do you do with the chip? At the WESCON trade show in September 1975, Chuck Peddle set up a table with two glass jars full of 6502 chips along with a sign: "$25 ". It was a radical change. The microprocessors of the day cost $200 or more. But the MOS chip fabrication plant could manufacture Bill Mensch's refined and simple design at incredibly low cost. Spot knocking allowed them to make over a hundred good chips from a single wafer of silicon, when Motorola could only produce ten. The MOS chip fab was an industry marvel ... right up until 1992, when the EPA designated it a superfund cleanup site due to leaks in their chemical storage tanks. Still ... it was just a chip. What did it do? To encourage customers to work with the 6502, Chuck Peddle embarked on a sales tour. He built a demonstration computer, the KIM-1, and produced manuals and sales materials. The board had one kilobyte of memory, which was enough to store a small program, read some input, and turn on or off the segments of an LED display. Business customers were convinced, and started ordering products from MOS; but as a computer it was not very exciting on its own. [1] The KIM-1 board (old-computers.com) At the same time, people who were not large corporations were equally fascinated. The Homebrew Computer Club, formed in the aftermath of the first computer hobbyist kits, provided the next spark: Steve Wozniak, the genius and polymath who designed the Apple I. Steve built his own 6502-based computer armed with nothing but ideas, encouragement, resourcefulness, a few loans, and even some funds from the sale of his personal graphing calculator. Steve Wozniak wanted something better than the calculator-derived KIM-1: the ability to see his output on a television. Since he did not have the backing of MOS technology, he had to use discrete logic -- individual logic chips connected together cleverly on the system board -- to create a television based typewriter output. [0] Actual output from an Apple I computer (YouTube) With their first two products generating excitement in the market, Chuck Peddle and Steve Wozniak both planned new products for 1977. A Tale of Two Computers Now that MOS Technology had been acquired by Commodore, Chuck Peddle's next computer would be named the Commodore PET. The PET was quite a hack: it was a KIM-1 with a television screen and a cassette tape player stuck on with duct tape. Yet that was how Commodore operated: engineers would hack together a product, present it at a trade show, and scramble to figure out how to build it if enough orders arrived. [0] The two 6502 based computers from 1977, from the HP9845 project The PET was shown in January 1977 and generated immediate demand; but it took until June 1977 before they were able to manufacture and sell them. As a business oriented computer, it only featured text, but enterprising developers figured out ways to play games nonetheless. [0] Gambling games on the PET, from Commodore.ca While the PET was a no-nonsense computer focused on business, the Apple II would instead be full of dazzling ingenuity and clever solutions, the kind that made Woz rightly famous. Wozniak was not just a computer designer: he also built arcade games for Atari. He created the arcade game Breakout, a one-player game using mechanics similar to Pong where you carved your way upwards. This experience led him to design the Apple II to support games. Both the PET and the Apple II could display a monochrome screen with 40 columns and 25 rows of text. But Woz had created the Apple II with graphics modes -- depending on how much memory you purchased. [0] Breakout for the Apple II, from Game Informer Graphics didn't come cheap. At launch, it would cost at least $1500 to buy a 12KB Apple II model capable of displaying "HIRES" 140x192 graphics. Even though it was possible to build computers with powerful graphic capabilities, it simply cost too much money to store graphics in memory at the time. The cost and performance of memory The base configurations of the PET and Apple II included four kilobytes of memory, about enough to store one page of a book. Still, four kilobytes of RAM cost a lot of money! One 1977 advertisement for a four kilobyte memory board showed a list price of $145 -- unassembled. How much memory did you need for graphics? The simplest way to store graphics is using a bitmap, where one bit of memory represents one pixel on screen. A monochrome 320x200 bitmapped screen would require eight kilobytes of memory; to display color would take double or quadruple that amount. Even if you could afford the memory required to store all the information necessary for a full screen of graphics, early 6502 chips took time to update the screen. Let's break this down: * The 6502 ran at a speed of about one million cycles per second. * A television refreshes its display 60 times per second. * The fastest data copying routine I can locate for the 6502 chip requires 30 cycles to set up and 16 cycles per byte copied. * For each frame on a television, a 6502 has around 16,000 cycles to execute code. Adding all those items together tells us that the Apple II and Commodore PET could, at best, move about one thousand bytes of data per television frame. So even if you could afford eight kilobytes of memory, you could only change part of the screen per frame. Early 6502 games tended to look like Oregon Trail or Lemonade Stand: they were slow, turn-based, and thoughtful, perhaps with a tiny animation or two. In fact, many games of the era completely dispensed with graphics and presented text-based screens (video, screenshots). A forty column text screen contained only one kilobyte of data, which meant that it could be updated in significantly less time, even if the art was less appealing. The next milestone would come from engineers who figured out a way to fuse color graphics and smooth animation while dealing with the cost of memory. The Atari VCS: Racing the Beam In the middle of the decade, Atari was well established in arcades and looking to expand. In 1975, it released the popular Pong arcade game as a home entertainment product sold through Sears. They knew there was a market for games at home. The advent of inexpensive microprocessors meant that it was now possible to design a game player, one that could swap cartridges in and out similar to videotapes or records. The founders of Atari invited Chuck Peddle to visit and show off the 6502 to Atari employees Joe Decuir and Jay Miner. The resulting product was so revolutionary that it didn't matter when other companies beat Atari to market. On September 11, 1977, Atari released a small device containing a cost-reduced 6502 and a Television Interface Adapter, also known as "Stella." The product sold for $199.99 -- a full one thousand and four hundred dollars less than an Apple II that could display graphics. [0] The $199.99 Atari Video Computer System, from Wikipedia The designers of the Atari VCS realized that they didn't need expensive memory chips to store graphics. The electron beam in a cathode ray tube moved at a slow and predictable rate. As long as programmers were careful, the 6502 and Stella could team up to produce a stream of colors on the fly and feed them to the television as the beam was moving. Fortunately, after the electron beam finishes drawing one frame, the beam is turned off and the television resets its magnets to point at the top left corner of the tube, so it can be ready to draw the next frame. This is called the vertical blanking process, and during the vertical blank period the 6502 chip inside the Atari VCS is free to think and compute game logic. [1] From "Stella Programmer's Guide" by Steve Wright All told, the Atari VCS was busy drawing pictures about 75% of the time. The remaining 25%, about 5,000 CPU cycles, could be used to update the game. Since most 6502 instructions took about 2-3 cycles, the programmer could expect to be able to execute about 2,000 instructions before it would be time to begin drawing the next frame. This wasn't a lot of computing power, but at the time, games were simple. And it had an interesting side effect: it was simply not possible for a VCS game to slow down. The game would either be able to display something on the screen, or it would display nothing. How did Atari graphics work? The team building the Stella chip generalized the ideas of Pong and refined them. They built the console and its first game, "Combat", together. [0] Atari Combat, the iconic game that drove the console's design The functionality needed to create Combat drove the design of the chip: there must be two "players" on screen. Each player should be able to fire a "missile". There could be a "ball" floating on the screen. And the "playfield" should be mirrored, so that each player had an identical home screen. One hundred and ninety two times for each frame, the 6502 chip in the Atari VCS provided information to Stella about the position, shape, and color of two players, two missiles, one ball, and one playfield. When the electron beam blanked out as the CRT turned itself off and returned to the top, the game had time to think and plan for the next frame. In the book Racing the Beam from MIT Press, Nick Montfort and Ian Bogost describe the incredible ingenuity of Atari VCS programmers. With only six elements on screen, programmers created a rich and vibrant ecosystem of Atari games, far beyond anything the original designers had intended. The clever work done to adapt Pong into Combat gave the Atari 2600 an amazing lifetime: it was to dominate retail gaming from 1978 until 1983. The C64: The world gets hit with a Jack Attack Not to be outdone, the founder and personal embodiment of Commodore, Jack Tramiel, continued his ferocious campaign. He had sold thousands of Commodore PET computers in Europe for immense amounts of money, essentially allowing Apple and Tandy to dominate the home computer market in the United States. Jack's incredible machinations are well documented in the book Commodore: A Company on the Edge, by Brian Bagnall. Flush with success from purchasing MOS technology, selling the 6502, and introducing the Commodore PET, Jack challenged his team to produce the next great thing. In the era when computers and game consoles were not clearly separated, and in a company where Jack Tramiel ruled with an iron fist, Commodore engineers were given little guidance. Talented engineers were hired and fired on a whim; and amazing products like the Commodore VIC-20 lived alongside duds like the Commodore Plus 4. Engineers who had the fortitude to stand up to Jack -- or to simply sneak around and do things without permission -- sometimes produced world shattering breakthroughs. When you combine Jack's relentless drive with Commodore's ability to produce its own chips, products could be designed, iterated upon, and built more cheaply than anywhere else. As such, Al Charpentier and Charles Winterble decided to invest time in creating a world-beating graphics chip. They would merge the Commodore PET's character-mapped display with the ability of the Atari VCS to inject sprites dynamically throughout the screen. The resulting chip, the VIC-II, was an evolutionary descendant of the original VIC chip and bore much similarity to Texas Instruments' TMS 9918. The two engineers presented the chip to Jack Tramiel. Depending on his whims, they could have been fired; but instead he instructed them to stick it in a box and sell it as a computer. With one condition, of course: Jack wanted 64KB of memory. [0] The iconic C64 reused design elements from previous Commodore computers What had changed? Jack, owning a semiconductor fabrication plant, knew that chips were becoming cheaper over time. He had established offices in Japan and negotiated favorable deals. He knew it was only a matter of time before memory was cheap, and he wanted to be the first company with a cheap computer for the masses. Introduced in August 1982, the Commodore 64 became the best selling (and by some accounts most popular) computer of all time. The Commodore 64 finished off the Atari VCS empire. Despite lurid stories about E.T. game cartridges buried in the desert, the Atari empire died for three main reasons: * First, parents preferred to buy their children a computer rather than a game system; * Second, because C64 games were better than Atari games; ... and: * Third, because C64 games could be copied -- illegally -- on floppy disks. [0] Mail Order Monsters, one of the author's favorite C64 games The VIC-II chip provided significantly more graphical power than the Atari 2600's Stella. Since the VIC-II could store all the information about a single frame of video in memory, the CPU could be fully available to programmers for the duration of the entire frame if they wished. Even though the C64 was still unable to transfer more than a kilobyte of data per frame, the character set for the C64 was fully programmable. Clever programmers could replace the ASCII character set with 8x8 background tiles and update them by writing "text" to the screen. As we saw before, tiled background graphics could be updated eight times faster than bitmaps. In addition to all that, C64 games could place eight sprites anywhere on screen. Even trickier, a clever hacker could display more than eight sprites by reprogramming them as soon as they were drawn to the screen. [0] Summer Games for the Commodore 64 (C64 Wiki) The Commodore empire was never again to reach these same heights. For all its breakthroughs, the C64 was actually a Frankenstein machine built from spare parts lying around within Commodore headquarters. The C64 was a mishmash: sure, it had the revolutionary VIC-II video chip. Sure, it had Bob Yannes' amazing SID sound chip. But it used the CPU, Basic, and "Kernal" of the Commodore PET. It used the I/O ports, floppy drive, and plastic case of the VIC 20. These features were ultimately to prevent the C64 architecture from outlasting this one machine, although Bil Herd's impressive C128 gave it a good try. Even worse, Jack Tramiel's famously aggressive attitude destroyed the Commodore name. He crushed Bill Gates and got Microsoft BASIC for a song; but had no plan to add graphics routines. The Commodore PET division had no clear architectural roadmap and was eventually buried by the IBM PC. Bill Mensch developed a successor to the 6502 called the 65816, but it was no longer owned by Commodore. Without a plan, talented engineers left in droves. The next year, Jack himself left Commodore after an argument with its board of directors. Commodore imploded. It never recovered, and the era of the 6502 began to wane. I've always been fascinated by the 6502 era. My father was heavily involved designing and manufacturing printers for early personal computers, and he brought home Commodore PET and C64 computers. During this crazy time, anyone could design a computer from scratch. Nobody knew how things were supposed to work. Titanic personalities roamed the earth. This included epic egos such as Chuck Peddle, Jack Tramiel, Steve Jobs, and Nolan Bushnell; and also quiet and unassuming technical geniuses like Steve Wozniak and the team that designed the original Macintosh. As this article hints, the 6502 was rapidly eclipsed by other processors. However, it did still power two more impressive games systems, both in Japan. We'll talk more about that in the next article in this series. -- -- -- More from tedspence.com Follow Software development management, focusing on analytics and effective programming techniques. Read more from tedspence.com Recommended from Medium Pacific Atrocities Education Pacific Atrocities Education Post WW2 Japan Reflected through Marie Kondo [0] The Institute of Geographical Studies The Institute of Geographical Studies in Geography ... everywhere! Sacred geography ... on the move! Jennifer R. Povey Jennifer R. Povey It's Black History Month -- Have Some Scientists [0] Archish Prakhya Archish Prakhya in The Academically Driven Symbol Of Hate? [1] Grant Taylor-Hill Grant Taylor-Hill in History of Yesterday Inchkeith: The Island of Plague, War, And Abandonment [1] Jennifer R. Povey Jennifer R. Povey The Time Pepsi Owned A Navy [0] Christopher Rodrigues Christopher Rodrigues Falling Together [1] Matador3364 Matador3364 Operation MissingNo, Part 4: The Deep Dive into Divinity [0] AboutHelpTermsPrivacy --------------------------------------------------------------------- Get the Medium app A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store Get started [ ] Ted Spence Ted Spence 1K Followers Software development management, focusing on analytics and effective programming techniques. Follow More from Medium Arpit Rathore Arpit Rathore Automating Maxima [1] Jake Cutter Jake Cutter in Technology Hits Computer Science -- Test Set & Clear in C++ [0] Raj Upadhyay Raj Upadhyay Introduction to Dynamic Programming on Grids: [1] Pradeep S G Pradeep S G Hello World in every programming languages [1] Help Status Writers Blog Careers Privacy Terms About Knowable