Sony’s success with PlayStation was already on the radar at Microsoft, causing emotions like concern and curiosity among the executives, but so far these responses had not quite escalated to great heights of fear and uncertainty. In mid-March 1999, with their preview of PlayStation 2 at the Game Developers Conference, Sony dropped a bombshell that represented a tangible threat to Microsoft’s game business. Where there was fear, there was opportunity, and the xBox team harnessed the PS2 boogieman to amp the threat level to Orange, which helped add urgency to the nascent discussions about Microsoft’s future game strategy. In doing so, they took advantage of particular Microsoft tendency—a tendency to assume the worst.
Brown and his co-conspirators understood this tendency. “One of the reasons we transitioned from a less aggressive approach to a more aggressive one is that we began to get into a more terrified state of, if Sony executes perfectly, what will happen? It’s not a bad way to think about your competitors… to think about what if? The beauty of that thinking is that you think about every possible, wonderful thing they could do. It can cause you to become a little bit over-analytical of what’s going on and to do some strange things, but there’s a balance point that sometimes Microsoft does really well. The early days of MS Explorer, early days of Windows, Office, early days of Xbox, there were these great points of paranoia and focus.”
The original PlayStation’s dominance of the gaming landscape meant that the console’s next generation was quickly seen as a threat to Microsoft’s growing PC game business. In fact, one Sony executive boasted that the PlayStation 2, with its dedicated gaming power and the proprietary Emotion Engine, would make PC gaming obsolete. The PlayStation 2 announcement also caused concern over something that was very important to Bill Gates. “It really was about the perceived threat that Sony was building an operating system for the living room,” says Ed Fries, “and Microsoft didn’t want to see that happen.”
Further fanning the flames, Berkes and Blackley coauthored a white paper entitled Evolution of Consumer 3D: The new Home of Visual Computing (See “Evolution of Consumer 3D” on page 249). The opening paragraph states:
“Recently, an uncomfortable fact has come to light. It comes not as a surprise—indeed it is rooted in the same commoditization of the PC space that has consumed so much of our bandwidth of late. In fact, the market and technological factors responsible have been obvious for some time. The problem, catalyzed by the announcement of the Sony PlayStation2, is a financial and technological oxymoron: the highest performance graphics machines in the world will soon also be among the least expensive.”
The paper included a brief primer on the evolution of 3D graphics, including the groundbreaking work of John Carmack, and the general drivers behind the relentless push for better and higher resolution graphics. But along with the “educational” elements of the paper were statements like:
…Sony has surmised that their next generation console, PlayStation2 (PSX2), must represent more than just a game box. It is, rather, an opportunity for Sony to actually own this new RT3D medium, which can generate far higher margins for them than any of their legacy media businesses. They also think that they can push the Windows PC out of this space, and keep it out.
The rest of the white paper details the rapid evolution of 3D hardware solutions, challenges presented by the comparatively slower evolution of CPUs, and solutions that have been found, such as adding memory and computing power directly onto graphics boards. At the end, Berkes and Blackley contrast the approaches that the PC and PlayStation 2 employed in implementing real-time 3D (RT3D), concluding that the more open PC model has many advantages over the closed console model adopted by Sony and other console manufacturers. In essence, they could be seen as making a case for their Windows-based console approach.
The Question of OEMs
In their internal discussions, the xBox team had discussed the living room strategy, recognizing it as a new competitive arena for Microsoft, which up until then had primarily focused on business or the home office. They were looking for a way for Microsoft to harness its investment in computer graphics and the DirectX APIs, as well as the developer “ecosystem” they had developed. How could they use their DirectX assets to compete with Sony for dominance in the living room? And how could the xBox team use Microsoft’s growing concern over Sony to their advantage?
“The perception was that everyone who played a game on a console was doing so at a risk to Windows,” observed Bachus, although, coming from the game industry, Bachus disagreed with that assessment. He believed that consoles and PCs coexisted because they did very different things, and even the games played on each platform tended to provide different types of experiences. Despite seeing no direct conflict between the two platforms, Bachus, along with Blackley, continued to support the idea of a Microsoft game console, even as unlikely as it seemed. Meanwhile, the unofficial xBox group continued to do research and refine their agreed-upon approach, which was still to enable gaming PCs built by OEMs, using a proprietary version of Windows specifically tailored to that purpose.
The more they considered the OEM approach, the more they realized how rigorous and specific the specifications would need to be, and how much support Microsoft would have to provide, not only in terms of technology, but also in design and manpower. They realized that they couldn’t realistically support very many OEMs and keep the program profitable financially. They started running numbers based on supporting five OEMs, but over time it got honed down to two… just two OEM companies to carry the Microsoft gaming logo and create these gaming machines to compete with Sony.
Another problem was the OEMs themselves. According to Brown, “I had visited these big OEMs and they weren’t designing shit. They were basically white boxing, white labeling everything out of China, and I’m thinking, you get 3% and all you’re doing is putting a sticker on it. Dell was still doing stuff in-house, but nobody else was. They were just trimming costs and using the cheapest labor in China. Anybody can play that game if they’ve got enough money for the other side of the game, which is marketing and sales and content.”
Despite their growing concerns over how they could implement a successful logo program that was profitable to all involved, they continued along the OEM path. Even as it became more and more logical for Microsoft to build its own console system, they still didn’t really believe that was a viable option. On the other hand, Bachus and Blackley, who say that their vision was always to build a Microsoft console system, saw the OEM solution as a sort of Trojan Horse—a smokescreen to keep the conversations going.
With further research they were able to extrapolate Sony’s potential dominance in key areas, and they made sure that Gates and the other execs were fully aware of their conclusions. For instance, PlayStation was able to produce an intense and immersive 3D experience, with a fast boot-up time, and even wireless controllers. (They didn’t actually have wireless controllers yet, but it was a threat.) On top of that, Sony manufactured a lot of TVs and monitors. It wasn’t too big a stretch to imagine what Sony was already implying, that the PlayStation could ultimately become a general computing device and take over the home. Moreover, the more successful consoles like PlayStation became, the more PC game developers might abandon Windows development, an eventuality that became increasingly important as more Microsoft execs came to understand the impact that games had on PC hardware sales. Hase’s group verified this fear by conducting surveys in stores, where they revealed that most consumers were purchasing PCs for the home to play games and surf the Web. And as consoles offered multiple uses, such as watching DVDs and streaming entertainment in addition to playing games, they became an even better sub-$500 investment for home consumers.
A Summit
Because the DRG was so deeply involved with Microsoft’s game technology, and especially due their role in developer support, they were often consulted on strategies that involved the game industry, and they often attended various planning and strategy meetings outside of those specific to the console debate. According to Bachus, it was at one such meeting in March 1999 that Gates asked them to plan a summit for May, inviting everyone at Microsoft whose interests touched on the game industry. Gates also asked for an analysis of PlayStation 2, especially in how it would compare graphically with where Windows was going. Based on their research, Hase and Blackley created a slide deck, with additional input from Bachus, who remembers, “One of the things that Seamus and Otto put into the slide deck was that we should totally make a console.” So, despite the general belief that Microsoft would never make a game console system, they didn’t back away from the idea, but sent the document, complete with the console recommendation, to Bill Gates days before the game strategy summit.
Brown explains that the idea of a “summit” was common at Microsoft. “Bill and other folks in upper management were always organizing or asking others to organize summits to get disparate and/or competing groups together to talk things out and/or demonstrate technology to execs or to one another. Summits were often used as a tool to defuse conflict in meetings and punt stuff down the road, sort of like the president appointing a commission to look into gun control or abortion rights and come back with recommendations six months later, by which time the anger or furor over the pressing issue might have passed a bit.”
The May summit included a lot of different groups from Microsoft—essentially anyone whose work might touch on games. In addition to the xBox team, Craig Mundie’s teams of ex-3DO/WebTV/WinCE people, Ed Fries’ Entertainment Business Unit, and Rick Thompson’s Hardware division were all represented.
Different approaches were presented at the summit—all of which had been presented in smaller meetings with Gates—including the idea of using Windows CE as a gaming architecture since it was already being incorporated into Sega’s Dreamcast. The idea of a reference standard of certification for PCs built by outside manufacturers was also still on the table. Another approach stemmed from the fact that a lot of devices, such as a TiVo for recording video or a console system like PlayStation, were essentially just computers. If they had any interface, it was specific to that device’s purpose, and not intended to serve in any other capacity. So why not consider a Microsoft-built PC with a Windows kernel, but nothing more? Such a machine could recognize a disc and start up a game from that disc without additional overhead or interface. It could cache some game assets to a hard drive to improve future loading of the game, save games in progress and provide other services, but it would still appear and function like a console—a Windows-based console. According to Bachus, they thought at the time that they could implement such a system and have it on the market by the end of 2000, when PlayStation 2 was expected to launch.