X-Box vapour triggers Nvidia stock frenzy
But does any of it make sense?
Nvidia's stock went wild yesterday on what seemed to be fairly conclusive news that the company had won the graphics gig for Microsoft's X-Box games console, which itself is tipped to be announced by Bill Gates at the Games Developer's Conference in San Jose this Friday. When Bill flips the switch the stock may well climb some more, and it'll likely be a shot in the arm for AMD's Athlon too - but er, why? The X-Box spec has been widely telegraphed, and the general consensus is that it's going to be based on an AMD Athlon 600MHz with Nvidia 3D graphics, DVD, hard disk, game controller and - presumably - an operating system. If something of this ilk were ready to ship today it might make some impact, but the ticklish problem is that the only software it could ship with is Millennium, due out around mid-year under the Windows ME tag. Microsoft is positioning ME as a consumer OS, and we note that determined spinners from Redmond have already started to describe it as "the ultimate operating system for gamers." This could be an X-Box hint, but if so, it's a desperate one. ME will be Windows 98 with knobs on and added hardware support, but it's difficult to see how adding it to what essentially sounds like a PC spec could produce a console, especially as Microsoft will be aiming ME at consumer PCs in general. How would it differ? Tight software integration with a specifically narrow set of hardware? That's about the only way it could, in the near future, and with this class of spec it would have to cost something similar to the cost of a reasonably good specification games PC, so for the moment there's a clear differentiation problem. From a design and software development perspective, targetting the product launch for Q3 2001, as has been rumoured more recently, makes a lot more sense, because in that timeframe it will be feasible to get the hardware designs together and - maybe - to get some OEMs ready to roll with machines based on them. It also gives Microsoft time to develop a serious console operating system based on Windows 2000. This of course is where anybody who's spent time looking over Microsoft's past roadmaps shouts fire. Microsoft OS roadmapping doesn't just work in dog or cat years - time passes at hamster speed or faster. So if we're talking about developing an operating system 18 months before it's due to ship, all sorts of ambitious things can be planned for it, and it's highly unlikely that anybody, right at the start of the roadmap, is going to be able to see what's really going to happen in 18 months time. They're hamsters, right? They don't live that long. But the trouble is that a tight, fast, Windows 2000-based consumer operating system is today just about as much of a non-trivial project as it was around a year ago, which was the last time it got cancelled, and Microsoft invented Millennium instead. So anybody thinking AMD and Nvidia are going to make big bucks out of an X-Box launched on the 2001 Christmas market should maybe ask themselves when was the last time Microsoft shipped a new operating system (not a service pack on steroids) to specification, on time? Or, for that matter, when was it that Microsoft last produced a viable hardware specification without input from its old partners at Chipzilla? But no doubt Bill will be able to address these questions on Friday. ®
Sponsored: IBM FlashSystem V9000 product guide