Forbes sees XBox 360 on track -- and we're still skeptical
Tags: Microsoft, Sony, Games, Gaming, XBox 360, Playstation 3, Marketing
Forbes.com cites a Piper Jaffray analyst stating that XBox 360's launch is on track for November delivery of about 1 million units in the US, 400,000 in Europe, and 200,000 in Japan. They also expect 15 games available at launch, with 35-40 games available by year end, and two price points of $299 and $399, where the $399 unit bundles the hard-drive option, an extra controller, and memory.
Our take: We stand by our belief that the schedule will be very tight. Why? Well, let's think about PowerPC chip production and platform manufacturing. This is a prediction of 1.6 million XBox 360 units. You can't produce those all in one month, so the production line will have to start soon if they are going to hit their dates. My bet is that manufacturing has to start about October 1 to ramp up for the Friday After Thanksgiving sales (Black Friday). I'd expect a ramp rate of about 100,000 the first month, 250,000 the second, and 500,000 the third. That leaves deliveries almost three-quarters of a million short. So start September 1. Oh, wait, the final release-to-master design still isn't done yet.... Oops.
But that's not the whole story. those boxes need chips too, 3GHz chips made by IBM in East Fishkill, NY. This is a tougher story, because IBM has not successfully produced 3 GHz PowerPC chips in quantity yet (if that were not the case, Steve Jobs wouldnt' be switching processor suppliers). The Fishkill plant has a max production capacity of about 1/2 million chips a day, but that's for a well-understood, debugged chip. Further, not all those chips are good (in fact, most of them are no good -- big chips have very poor production yields, often in the single digits) and few of them run at the highest speeds. If IBM can get 3-5% yields on 3GHz chips early in production, that's great, but actual numbers will likely be considerably less. So expect IBM to be cranking out only about 30,000 Xbox capable chips a month for a while. And each XBox requires three of them.
The bottom line: I expect IBM to be the biggest bottleneck because of the unknowns in a new high-speed chip -- those production lines just take time to ramp up. And because XBox's design requires so many of the chips, I would be very surprised if XBox launches with anywhere near the number of units that Piper Jaffrey claims; my bet is that there will be at most about 1/2 million made, and that's only if everything goes perfectly (and we know they never do). Expect XBox 360s to be about as plentiful as hen's teeth at Christmas -- if they are available at all.
Forbes.com cites a Piper Jaffray analyst stating that XBox 360's launch is on track for November delivery of about 1 million units in the US, 400,000 in Europe, and 200,000 in Japan. They also expect 15 games available at launch, with 35-40 games available by year end, and two price points of $299 and $399, where the $399 unit bundles the hard-drive option, an extra controller, and memory.
Our take: We stand by our belief that the schedule will be very tight. Why? Well, let's think about PowerPC chip production and platform manufacturing. This is a prediction of 1.6 million XBox 360 units. You can't produce those all in one month, so the production line will have to start soon if they are going to hit their dates. My bet is that manufacturing has to start about October 1 to ramp up for the Friday After Thanksgiving sales (Black Friday). I'd expect a ramp rate of about 100,000 the first month, 250,000 the second, and 500,000 the third. That leaves deliveries almost three-quarters of a million short. So start September 1. Oh, wait, the final release-to-master design still isn't done yet.... Oops.
But that's not the whole story. those boxes need chips too, 3GHz chips made by IBM in East Fishkill, NY. This is a tougher story, because IBM has not successfully produced 3 GHz PowerPC chips in quantity yet (if that were not the case, Steve Jobs wouldnt' be switching processor suppliers). The Fishkill plant has a max production capacity of about 1/2 million chips a day, but that's for a well-understood, debugged chip. Further, not all those chips are good (in fact, most of them are no good -- big chips have very poor production yields, often in the single digits) and few of them run at the highest speeds. If IBM can get 3-5% yields on 3GHz chips early in production, that's great, but actual numbers will likely be considerably less. So expect IBM to be cranking out only about 30,000 Xbox capable chips a month for a while. And each XBox requires three of them.
The bottom line: I expect IBM to be the biggest bottleneck because of the unknowns in a new high-speed chip -- those production lines just take time to ramp up. And because XBox's design requires so many of the chips, I would be very surprised if XBox launches with anywhere near the number of units that Piper Jaffrey claims; my bet is that there will be at most about 1/2 million made, and that's only if everything goes perfectly (and we know they never do). Expect XBox 360s to be about as plentiful as hen's teeth at Christmas -- if they are available at all.
posted by Carl Howe at
11:03 AM
| View blog reactions
| 
