The History of Sega Console Hardware
Hideki Sato – 1998 Developer Interview
If I had to sum up succinctly what makes the Dreamcast special, I would say it’s connectivity. As you know, every unit comes equipped with a modem. However, this isn’t our first experiment with using modems. It’s actually our third try (after the Mega Modem for the Genesis and the Sega Saturn modem).
I’ve been told there are many Sega fans in Japan alone, and I’m very thankful for that fact, though I sometimes wonder if they aren’t just rooting for the underdog. (laughs) Or perhaps they are drawn to Sega’s forward-thinking posture, as we have always tried to stay on the cutting edge of technology. I can only guess at the exact nature of our fan’s mentality, but in any event, I am truly grateful.
The Dreamcast, therefore, is a challenge to us: to take our already wide userbase, and see if we can enlarge it further. In that sense, the Dreamcast represents a recapitulation for us of 20 years of determination, and a commitment to challenge ourselves and move forward in the home video game console market.
The First Challenge
Hideki Sato, longtime Sega company man and director of R&D during Sega’s 16-bit heyday.
From the beginning, Sega’s home console development has always been influenced by our arcade development. Our first 8-bit machine was the SC-3000. This was a PC for beginner-level users. At that time, Sega only did arcade games, so this was our first challenge. We had no idea how many units we’d sell. In our commercials we used the comic duo “Tunnels” to advertise the console—before it was even available on-sale! Including the overseas markets, we ended up selling several tens of thousands of units. Those were heady days.
We learned that Nintendo planned to release their own console system at the same time, so we removed the keyboard from the SC-3000 and marketed it as a new home console, the SG-1000.
But no one predicted how big of a hit the SG-1000 would be: the first year alone, we sold 160,000 units. They sold so well, we started to get stars in our eyes, thinking we might sell half a million! It was a scale completely incomparable with our arcade board sales. And that is how our entire company caught Console Fever, and threw ourselves into that side of game development. (laughs)
The problem was, while we knew how to make arcade games, we didn’t really know anything about console development. In fact, the very idea of a “consumer” market for video games was unheard of then: back then it was just a “new business” idea. We were lacking adequate staff to develop the software for our new system, too. Consequently, I don’t think we put out very good games back then.
Still, I think there were a few “Sega innovation” moments, even then. The best example would be the Sega Card media format. In-line with the “smaller, sleeker, faster” technology trends of the time, we aimed for an easily portable format. Also, we knew that many children liked to take their games over to their friend’s houses to play, so we tried to design something that would look cool and stylish when you whipped it out of your pocket.
SC-3000 tv ad, featuring an early celebrity video game endorsement from Japanese comedy duo “Tunnels.”
Those considerations led us to develop the Sega My Card: a game format with all the data chips in a space 2mm thick. However, we ultimately went back to a cartridge format, because games started needing more memory, and cartridges could provide more. But nowadays, when I see things like “smart cards” equipped with actual CPUs, I can’t help thinking how Sega was 10 years too early in that market. (laughs) Either way, the My Card is a particularly memorable moment in Sega history for me.
In the meantime, our development teams had been working on a new home console system, based off the arcade System II PCBs, which would reproduce that arcade graphics engine as closely as possible. This was the Sega Mark III. The Master System, which we developed specifically for the American market, was based on the Mark III.
However, Nintendo’s Famicom enjoyed a virtual monopoly on the domestic games market in Japan. It was similar overseas, with the Master System coming out an entire year after the NES. No matter how superior our graphics were to the Famicom, they had the clear advantage by virtue of being there first. Ultimately, we were only able to obtain about a 10% share of the domestic console market.
The 16-bit CPU
At that point, we decided to start developing a new home console. By then, arcade games were using 16-bit CPUs. Arcade development was something we were very invested in, so we were always using the most cutting-edge technology there. Naturally, it started us thinking: what if we used that technology in a home console? Two years after we started development, it was done: a 16-bit CPU home console, the Megadrive. The 68000 chip had also recently come down in price, so the timing was right.
We had a feeling that before long, consumers would be appreciating video games with the same sense with which they enjoyed music; moreover, since the Megadrive was a machine that you put in front of your TV, our concept was to make it look like an audio player. So we painted the body black and put the “16BIT” lettering in a gold print. That gold printing, by the way, was very expensive. (laughs) But we really wanted to play up the fact that this was the very first 16-bit home console.
The following month after the Megadrive console’s release, we put out Altered Beast. I remember very fondly a demonstration we did at an announcement party for the Megadrive, where we took the Megadrive version of Altered Beast and played it next to the arcade version, and there was almost no difference.
There was a certain self-conceit and pride at Sega back then, and we did almost all of our own software development. We didn’t rely on licensees much at all. I remember, at that time, seeing the first third-party developed game, Thunder Force II, and being surprised at how well it took advantage of the Megadrive’s capabilities. I suspect the other developers at Sega felt the same way. That initial high level of quality was a real motivator for everyone, and after that, more and more good Megadrive games started coming out.
One example of Sega’s ad campaign
comparing Nintendo and Sega hardware.
Following on these successes, we decided to release the Megadrive in America. The American version was called the Genesis. It was a big hit. With Sonic the Hedgehog and our commercials comparing Sega and Nintendo, we did a good job building the Sega brand in America. We sold over 15 million systems. The success of the Genesis in America really helped shape this as one of Sega’s strongest eras. We also received a reward from Motorola at this time, which stands out as another happy memory.
The Eve of the Saturn
While the Megadrive was a hit in America, in Japan, it was quickly overshadowed by the Super Famicom. In our usual way, we took this setback as a challenge, and released several peripherals for the Megadrive: the Mega Adapter, the Mega CD, and the Mega Modem. If you had them all hooked up, it was an amazing sight to behold! It felt like you’d put together some monstrous tank. (laughs)
The Mega Modem was our response to the recent developments in networking technology. At the time, PC networking was just starting to gather popularity. The baud rate then was 1200 bps. We used that rate for competitive baseball, mahjong, and similar games, but the level of technology made it rough. Moreover, we made very little money off the Mega Modem, so even at Sega, hardly anyone understood it. But from that experience we learned that networking capabilities had a lot of potential, and we resolved to include them in our next console. Sega was an “arcade game culture” company, you see, so we were always quick to get back on our feet. (laughs) In the arcade industry, just sitting back and waiting for the technology to ripen was never an option.
Advertisement for the Sega AI computer (click to expand). The tagline reads: “The computer that touches back. Starting today, you’ve got a partner in AI.“
The TeraDrive was our first multimedia PC, co-produced with IBM. At that time the word “multimedia” didn’t yet exist, though, so we had a hard time explaining the merits of this PC/Megadrive hybrid.
The AI Computer, on the other hand, was educational hardware. We originally intended the AI Computer to come equipped with a cassette tape deck, for re-writeable media. We figured it would be needed for the educational software, such as English learning programs. Unfortunately, the tape media was really awful: it had horrible loading times and wasn’t very reliable. We ended up with a hybridized system, where the voice data would be stored on cassette tape, but the program and other data would all be handled by My Cards. There was also some software that only ran on My Cards, if I recall. We mainly sold the AI Computer through mail-order and direct sales, and I think we only ever sold 5000 units. (bitter laugh)
Our previous experiments with My Cards and drawing tablets led us to create the Sega Pico. The tablet’s positional sensor technology came from technology we had developed earlier for our 1987 arcade medal game, World Derby. The graphics chip and CPU, on the other hand, were from the Megadrive. In a sense, the Pico can be described as a Megadrive+tablet+voice synthesis device. (laughs) Sega released a lot of different hardware, but they were all connected to each other.
After that, Sega set out to develop a new 32-bit console. We were then in the process of researching what kind of games we could create using CG technology. Driving games, flight simulators, but what else…? Yu Suzuki was working on an experimental project to render and animate humans in 3D CG. Eventually he showed us what he had been working on, and we all saw something we had previously thought impossible: realistic human figures rendered in CG. Once we saw that, we were no longer worried about the kinds of games we’d be able to make with this new technology.
At first, we couldn’t decide what media format to use for our new 32-bit system: cartridge or CD-ROM. Internally, we called the cartridge version Jupiter, while the CD-ROM version was called Saturn. We worked on both systems in parallel until midway through, when the larger storage capacity of CD-ROMs won the day for the Saturn.
Sega staff working on Golden Axe II. The designers are drawing sprites with Sega’s “Digitizer Mark-III”, a Wacom-style drawing tablet, likely the same technology pioneered in World Derby and used in the Sega Pico, as described above.
We were also stuck on whether to focus game development on sprite-based games, or new 3D CG games. Sprite-based games were what Sega had done up till then, so we had a lot of built-up experience there, both in a personnel and technology sense; it seemed like a waste to just throw it all away. And Sega only had a few internal teams dabbling in CG design. We therefore decided to give the Saturn the ability to handle both kinds of games, with a robust sprite and CG engine. However, although we’d separated the two engines well enough in a hardware sense, creating games for the Saturn turned out to be a little difficult. The software development libraries were also insufficient, so third parties saw the Saturn as a difficult system to develop for. We sold 5 million systems in Japan, but we struggled in the overseas market.
The Birth of the Dreamcast
Finally, we’ve reached the start of the Dreamcast development. This machine has a Hitachi SH-4 CPU. The Saturn also used a Hitachi CPU, the SH-2, so it was a fairly easy transition. As for the media format, we researched the possibility of using DVDs, but the costs were still too high. DVD movies are very popular right now, but those are comparatively easier to produce, technology-wise. If we were to use that media for games, we’d have to research authoring issues as well. All that would have taken too much time, so we did not adopt DVDs.
Our next idea was to include a rewriteable CD-ROM drive. We didn’t go with that idea, but it ultimately led us to adopt a new technology, the GD-ROM, which was a natural extension of the CD-ROM’s technology, with double the storage capacity.
Finally, there was the graphics engine. We ended up with two contenders: the American 3dfx’s chip, and NEC’s Power VR2. The 3dfx chip was of an orthodox, typical construction, while the Power VR2 showed the way to the next generation of graphics. At the end of a long series of debates and comparisons, we chose the Power VR2. It’s high quality graphics are 100 times more powerful than the Sega Saturn and other previous consoles. Furthermore, NEC also plans help support Dreamcast development by incorporating the same graphics chip in its own line of personal computers. The collaboration with Microsoft and Windows CE was likewise done for the same purpose, to make game development easier.
The Dreamcast development keyword, as described
by Sato below, was “play and communication”.
In Search of New “Play”
With the Dreamcast’s specs and media format decided, we came to the question: what next? With our previous consoles, we would have stopped right there, but for this new machine, in order to really be satisfied at Sega, we knew we needed something special to distinguish our console from others. It was here that we first considered the option of adding a modem as a standard feature.
On the Saturn, we had the Saturn modem and the XBAND online service, with about 15,000 users. When we looked at how they were using the XBAND service, we discovered that the usage was split about 50/50 between vs. fighting match-ups, on the one hand, and e-mail on the other. To be honest, there weren’t really a lot of networked games available to begin with, so I’m not sure how meaningful the data about the vs. fighting games was, but the part about the e-mail showed us that that this was a significant user need. From there we asked, could we capitalize on this and perhaps simplify e-mail, and find new kinds of “play” based around that need? Could we develop a virtual space for networked communications, but give it a “game” feel? We want to create a community for Sega users to communicate with each other: we think a new kind of gameplay could emerge from such a space.
So it may be the modem that is the key to all this, to broadening the horizons of games. With graphics and sounds, if you don’t increase the power of a new console by a magnitude of x100, the average user won’t really notice the change. That’s why you have to find some new direction, some new angle, when you create a new console. The modem represents that new direction we are presenting to players with the Dreamcast.
Hideki Sato breaks it down.
I think it may be enough just to create an online space for players, and the rest will sort itself out. That may only be a slight exaggeration. Right now, what people value in games is diversifying. It’s extremely hard, therefore, to design a single game that will please everyone.
Take, for example, the Print Club (puri kura) photobooth kiosks. High school students have created new ways of interacting and playing with those machines, ways the original creators never intended them to be used. Then, through mass communication, the phenomenon spread even wider.
Today more and more players are finding value in that kind of game: where the developers need not specify the exact kind of “play”, but can leave it open for the users to freely create themselves.
To our thinking, online networking contains the same possibilities. People always want to communicate with others. When you look at the content of most actual internet home pages, that’s the goal. It’s a cool feeling, to have anyone capable of sharing information and experiences like that. With the Dreamcast, Sega wanted to create more of those spaces.
Part II: Console by Console
After having spent many years working in the arcade business, Sega jumped into the console market with its first system, the SG-1000. I was very lucky to be working at Sega at this time, and to be a part of the planning for this console. As you can imagine, in the beginning we were shocked at the difference in sales volume between the arcade and consumer markets. Back then, an arcade machine of ours would only sell about 5000 units, so even a small change in production costs could yield a good profit (if ICs went down 10 yen, for instance, we’d make 50,000 yen). However, take that same price differential and apply it to the console market, where we were moving upwards of 300,000 units, and suddenly that 10 yen savings nets 3,000,000 yen!
For that reason, we really agonized over the exact choice of components. We estimated that in a single year, the average user would use our consoles for about 500 hours, and we based our choice of chips and their durability/lifespans on that estimate. Using low quality parts is obviously a bad idea, but using parts that are too high quality also isn’t good. As a result of that scrutiny, though, we were able to sell the SG-1000 for half of what the SC-3000 cost, and in our first year we exceeded our sales estimate by selling 160,000 systems.
The SG-1000 and its more stylish redesign, the SG-1000 II.
The 1980s was the decade when the “information society” took shape in Japan, and Sega’s contribution was the SC-3000 computer. We released a BASIC programming platform and educational software, but when we took a closer look at how people were using the machine, we discovered that half of the users only played games on it! At that point we swapped the keyboard out for a joystick, and sold the SC-3000 for half-price as a specialized game machine, the SG-1000. The CPU was an uPD780-1, which was compatible with the MSX z80A chip. Looking at the market around us, we figured we’d sell about 50,000 systems, but we actually sold three times that.
However, now that time has passed I can say that the design of the SG-1000 was, in fact, really horrible. When you inserted one of those big old cartridges into the system, the whole thing looked like some big tombstone… (laughs) But the good sales invigorated us, and for the next year, we redesigned the exterior of the SG-1000 to look cooler, a “mark II” version… to wit: the SG-1000 II.
Sega Mark III
Since the early days of the company, Sega had employed a lot of able developers, so we always placed a heavy emphasis on doing our own in-house software development. This posture continued up to the 32-bit era; however, it also meant that we never really cultivated strong relationships with third party developers. Even in the 8-bit era, our attitude was “we’ll just port our arcade games to our consoles.” And so we did, but it resulted in a large difference between our game library and that of Nintendo’s Famicom.
Nintendo had courted third parties from the start, and from 1984 onwards the gap in sales between the SG-1000 and the Famicom only continued to grow. The following year, we saw that we needed a “mark III” of the SG-1000 to compete, and we decided to change the graphics chip. The TMS-9918 we had been using was fine for simple computer games, but in truth it was simply lacking in power for the kinds of games we really wanted to make, with more (and more beautiful) graphics. As such, we resolved to create a new graphics chip ourselves, and the system we built around that was the Sega Mark III.
The Master System was especially successful in Europe. This gave us a foothold in the overseas console market, and so we started thinking about our next console. In the arcades, Sega had long been using the system board format for our arcade games. At the time we had the 8-bit System I and System II boards, and we had just finished developing a new 16-bit board based on the 68000 chip, the System 16. We wanted to use the same technology for the home console market too, a 68000-chip system with arcade-level graphics. To this we added FM sound, and the Megadrive was born.
The legendary 68000 CPU (notice the S for Signetics), which
was very nearly a casualty of business negotiations.
At the time, a 16-bit home console was unheard of—the idea itself was sensational. Now that some time has passed, I can tell you a story of how the Megadrive came to use that 16-bit 68000 chip, and how it almost didn’t happen. Back then, we were involved in negotiations with a company named Signetics, who had a license from Motorola to manufacture 68000 chips. Whether we used the 68000 or not depended on how these negotiations went, and we were asking for a price that we knew was unreasonably low. But when we finally concluded the negotiations and successfully obtained a blanket order of 300,000 chips, everyone was elated. “We got it!!!” Selling that many consoles was going to be tough, but it felt like the bigger battle was behind us.
The design of the console went through refinement after refinement, until we arrived at a black body with the “16-bit” lettering in shining gold print. It was the height of the console “bit wars”, and bigger numbers gave an impression to consumers of more power. We made the reset button blue, and we wrote the phrase “INTELLIGENT TERMINAL HIGH GRADE MULTIPURPOSE USE” in red lettering, which wrapped around and framed the cartridge slot. We wracked our brains to come up with that sentence, but I think all our hard work paid off, and we were rewarded with a machine that looked very stylish.
When the Megadrive was first released, some of the games were pretty lousy, like Osomatsu-kun… but when it came to visuals, the Megadrive was unrivaled by other consoles. And before long, games from third party developers started coming out that really made use of the hardware. It was a reaffirmation for us, that we were not mistaken in our design philosophy for the Megadrive: “a high quality product from limited source materials.” Many great developers got their start with those early third-party Megadrive games, too.
With the NES having beat us to the punch, the Master System didn’t have much hope of gaining market share overseas. Accordingly, we set our eyes on being the first to bring a brand new console to the American and European markets. In 1989 we introduced the 16-bit Genesis to America, but our competition was no longer limited to Nintendo: NEC had also released their own very capable system to the American market, the “TurboGrafx 16”, as it was known overseas. Even though it was technically an 8-bit machine, it was mistook for a 16-bit system, adding more fuel to the fire of the “16-bit wars” between the Genesis and SNES.
The Megadrive plus peripherals.
Or, as Hideki Sato describes it, a “monstrous tank” !
Really, these “bit wars” continued in some way up to the Dreamcast, because consumers had the impression that if the numbers doubled, then the power of the machine would be doubled too. That, of course, is why we put “16-bit” so prominently on the Megadrive. In Japan we used sparkly gold lettering, but in America we were told that the gold would be easily mis-seen as yellow, which wasn’t a very popular color. So the “16-bit” became silver colored, and the reset button was changed from blue to grey for similar reasons. We made the “16-bit” lettering a little smaller too, aiming for a smarter looking design. These design changes reflected the difference in culture and values of the American and Japanese consumer. Thanks to those efforts, and the success of Sonic, we were able to compete on equal footing with Nintendo during this era.
Nintendo’s Game Boy was black and white. At Sega, we saw that as a challenge to make our own color handheld system. Again, though, I think we may have been ahead of our time. (laughs) As for the hardware design, we began by calculating the weight of the batteries the device would need. We worked backwards, adding up the weight of every component: batteries, backlight, cartridges, to arrive at an estimate of the total weight of the system.
At the time, the Game Boy weighed 270 grams, while Atari’s Lynx was about 800 grams. So we aimed for ~500 grams for the Game Gear. Overseas, a converter was sold which allowed the Game Gear to play Master System games, and if you add in the domestic sales, we sold about 14 million systems: a respectable chunk of market share. However, as you know, Nintendo’s Game Boy was such a runaway success, and had gobbled up so much of the market, that our success was still seen as a failure, which I think is a shame.
The Game Gear + TV Tuner + Wide Gear (a screen magnifier).
Truly the aesthetic brethren of the Megadrive!
When the Saturn was being designed, the video game industry was right in the middle of the transition from sprites to CG. In the arcade world, you could see this contrast between the System 32 boards, capable of displaying 300,000 sprites, and the Model 1 boards which ran Virtua Fighter and showed the future of polygons. In order to not lose all the assets and know-how we’d accumulated in previous years, we first thought about basing the Saturn on the System 32 boards, but we inevitably realized that it would be best to have polygon and CG capabilities too, so we included both in our design. It was done in the spirit of having the best of both worlds, but it also kind of felt like we were splitting the baby, and not doing justice to either. (laughs)
There were two candidates for the CPU. The first, which Sega of America was pushing for, was the 68020. It had good compatibility with the 68000 processor and would be easy to use, but its limitations were also clear. The other option was the RISC CPU: it seemed much more powerful, but for several reasons, the risk was also much higher (just as the name “RISC” implies!). As it had always been with Sega, we needed a home console that would be powerful enough to handle our arcade ports. That being the case, we took the risky-but-idealistic path and selected the RISC processor, the Hitachi SH2.
The keyword for the Dreamcast development was “play and communication.” The ultimate form of communication is a direct connection with another, and we included the modem and the linkable VMUs for that purpose. We had also planned to have some sort of linking function with cell phones, but we weren’t able to realize it. Consumers were now used to the raging “bit wars”, so even though we knew it was a lot of nonsense, we needed to appeal to them in those terms with the Dreamcast. And so we marketed it as having a “128 bit graphics engine RISC CPU”, even the SH-4 was only 64-bit. (laughs) On the other hand, we extensively customized the original SH-4 for the Dreamcast, to the point where I think you could almost call it something new.
The modem in Japan is 33.6kbps, and in America it’s 56kpbs, but we designed the Dreamcast’s modem to be removable and upgradeable with advances in hardware and infrastructure. This cost a lot, but we were thinking about the future. We also had the idea of making the GD-ROM drive front-loading and even made several prototypes in that style. Personally, I wanted a more cube-shaped design like the Nintendo Gamecube, but we ultimately ended up with what you see today. One thing I truly wanted was wireless controllers. Those long cords are annoying. Next time, it will be wireless for sure! (laughs)
Hideki Sato signs a Megadrive—with a gold sharpie, of course.