This article, replicated on TeamXbox and several other Xbox sites, is a "leak" of plans that Microsoft apparently has for the next Xbox. In this article I will deliberately be inflammatory about this issue -- as an owner of two current Xbox systems, a proud Xbox Live subscriber, and owner of close to twenty games, I feel this gives me some right to speak on the issue.
I know that these plans are neither official nor likely, because of the ridiculous suggestions presented in them, but here's one gamer's opinion on what will happen if Microsoft goes this way. Note that many of my answers refer to Dean Takahashi's book, "Opening the Xbox".
Three IBM-designed 64-bit microprocessors.
Idiocy. The learning curve for developers to take advantage of simultaneous processor systems is "horrendous" and can range from twelve to eighteen months. Even on current-generation consoles with one processor, we still get shovelware, portware, and poorly-executed user interfaces. Two of these IBM CPU's would be sufficient, much like in the Mac G5; one thread can handle physics calculations, while the other can concentrate on providing data to the GPU, for example. Developers are getting used to dual-processor systems; don't push the envelope. This goes against Microsoft's first strategy when creating the Xbox, which was to make programming PC games and programming Xbox games nearly the same process.
Having said that, I believe the G5, or a processor similar to it, would be one of the most ideal solutions for the third-generation consoles scheduled to arrive in 2005/2006. My preferred idea would be similar to a high-end gaming PC, where AMD supplies a 64-bit CPU with gobs of speed. A 2500+ chip retails for about $135CAD nowadays. By comparison, the closest Celeron chip to this price point is the 2.4GHz, or a Pentium III running at 1GHz. One of the lowest end Pentium 4 chips - a 1.7GHz with 256KB of cache -- starts at $215.64CAD.
A graphics chip designed by ATI Technologies with speeds much faster than its upcoming R400 chip for the personal computer.
This is a good thing. ATI's cards are rock solid performers now, and their chips consistently have a lead on the newest nVidia models -- the 9700 Pro was the undisputed performance king for about six months while nVidia gawked and frittered around with their FX line of cards. The 9700 Pro is still a damn good card and if I didn't already own a 9800 Pro, having a newest-generation ATI chip in a console would make said console pretty cool.
Compatibility with the original Xbox, which is based on Intel and Nvidia chips, isn't guaranteed.
Completely lousy business decision. Why does Sony have more consoles out in the market? I would WANT backwards compatibility and likely would not buy the console at launch without it. I especially would dislike switching between consoles, and would end up holding out for a grey-market emulation solution like Cxbx so I could emulate the games on my PC. Microsoft depends on the people that absolutely must have the console at launch -- that basically pay a premium
for having it a few months in advance. At $399CAD for the original Xbox, I didn't even consider it. For $299CAD, I sprung for one because of word-of-mouth. And for $249CAD, I bought a second unit for online play. You have to reach the price point with gamers.
In contrast with the current Xbox, the next one will have no hard disk drive -- unless Sony puts one in the PlayStation 3.
So you expect me to go out and pay $40 to save my games, with the promise of zero things that made the Xbox great -- ripping CD's to the hard drive, precaching levels such as Halo does, upgradeable dashboards when online experiences change, downloadable content bigger than the size of a memory card? Oh, and possibly Microsoft might sell an additional hard drive to attach to the console, which you can't be guaranteed will work. The only way this concept might conceivably fly would be extremely large memory cards -- somewhere in the area of 1GB or greater. KOTOR savegame files can get up to 2000 "blocks", which is about 32MB per savegame.
The machine also will have about 256 megabytes of dynamic random access memory.
512MB is more reasonable for intensive gaming. 1GB might be pushing the envelope, but whatever happened to making the greatest gaming console ever? Guess that vision died a while back when someone considered dropping the hard drive. Can anyone give me insight on how Halo Mac performs on a G5 with 256MB RAM, such as the lowest costing model on Apple's site?
And lastly, it isn't clear if Microsoft will include the current DVD video technology or Blu-Ray, its successor.
Standard DVD-ROM media still seems to be the way to go. Games for PC haven't started shipping exclusively on DVD yet, and games like Halo PC fit on a 650MB CD-ROM. If Microsoft does do one thing right with the DVD drive, though, don't choose Thomson as your manufacturer, and go exclusively Samsung for the next
Xbox units. Not only are the drives more reliable and read more types of CD media, but they also tend not to die in a year.
For the sake of people who actually like your existing console, Microsoft -- you know, the four million people that bought Halo for Xbox? Don't screw this one up, or all the smart people will stay with PC gaming.
***
Addition to Article February 01, 2004: I notice that the author of the Mercury News article is none other than Dean Takahashi himself. If you're at all interested in how Microsoft got around to being involved in the console business, "Opening the Xbox" is the definitive book with the answers. This also lends a bit of unforseen authenticity to the news; which makes me think, "What is Microsoft doing?"
