AMD's announcement of four new mobile processors gives us a glimpse of what the rest of the year will be like. AMD has said repeated times last year that their goal is 50% or more of the mobile PC market. They also said that all new "net capacity increases" will go to mobile; this will be several million additional processors per quarter -- a significant portion of the 7 million per quarter mobile processor market. If they can continue to broaden their product line (add T&L and SFF) and continue to increase performance, I think roughly 50% mobile market share by year-end is possible, though they may lose desktop share to get there. Become a Complete Fool
I've been thinking about what this may mean for AMD. In general, it probably means ASPs are going up.
Let's take a look at AMD's latest price lists along with an estimate of the average selling price (ASP) for each part (assuming a 40% discount to list).
Current AMD XP Mobile Processor Pricing for 15W-25W range
Processor List Price Est. ASP
Athlon XP Mobile 1400+ $190 $114
Athlon XP Mobile 1500+ $250 $150
Athlon XP Mobile 1600+ $380 $228
Athlon XP Mobile 1700+ $489 $293
Assuming a normal distribution of bin splits, that's a weighted ASP of around $190. Not bad at all compared to the current overall average ASP of $90. So if all new production gets added to mobile, ASP could increase a lot -- especially as the Duron is ending production in Q2 and ending shipments probably in Q3 or Q4.
If AMD keeps introducing faster mobile parts, AMD can keep a high mobile ASP the rest of the year just introducing higher performance mobile processors.
However, that's not the full story. AMD is also releasing thin and light processors later in this quarter. At a minimum, thin and light notebooks need smaller processor packages than full-size notebooks do. They also typically have lower clock frequencies in order to reduce power requirements so that smaller batteries still provide adequate runtime on a full charge.
According to some of the slides at the analyst conference, AMD plans to release processors in the under-15W range for thin-and-light applications with the new micro-PGA packaging. This should allow AMD to sell high-ASP parts at low voltages and low power that may not bin at the highest performance level.
Intel already appears to be taking advantage of the higher value of T&L. Note that Intel sells 850 MHz PIII-Ms as T&L processors for over $300, even though these processors probably could not be clocked to 1.26 GHz and sold as highend PIIIs at full power for $200. From the same press release, you can see that Intel charges $144 for a 650 MHz ultra-low-voltage (ULV) Celeron, even though a 1.06 GHz low-voltage (LV) Celeron is only worth $106. Clearly lower power parts even at much lower clock rates sell for more.
Lower-power and T&L is worth a premium to Intel mobile customers -- as much as a 50% price premium. I doubt T&L versions of mobile processors will add that much of a premium for AMD, but I do expect AMD will be able to charge 10%-15% more on their T&L parts than their standard mobile parts for considerably lower power.
I expect AMD to be able to release T&L parts in the 1200+ to 1300+ range within a quarter that have 15 W maximum power (and under 1W typical power like Intel's LV parts). They don't seem to be going after the ULV market, but that is a much smaller market anyway; T&L is where most of the market is heading -- there is not much volume yet in the ultra-portable market.
Intel doesn't have much room to respond. It's unlikely that Intel can release 1 GHz T&L processors with under 15W max power based on the PIII design, since they are having trouble clocking the desktop PIII much higher than 1.3 GHz at full power. Intel is also unlikely to release a 1.2 GHz P4 with less than 15 W max power; their current P4s seem to be in the 35W max power range at 1.6 GHz. It will take a lot of work to get the P4 down to lower power. Intel probably has to wait for Banias in 2003 to have a superior or equivalent mobile solution in most subsegments.
It looks to me like AMD is going to be able to match Intel's GHz with equivalent model numbers in 25 W or less range mobile processors while having 10W less power dissipation than the mobile P4 and be able to exceed Intel's mobile PIII clock rates at LV power levels. Intel will have to compete at a disadvantage in mobile in the biggest mobile segments. I expect AMD to have such an advantage in standard and T&L to be able to price some of their mobile parts higher by year-end than Intel does. Already, AMD mobile pricing seems to be inching up on Intel.
Will Intel respond by cutting mobile ASPs to desktop levels to prevent market share loss or will they let AMD make higher profits than Intel while their own pricing power diminishes in some parts of the market? I can't really say, because I don't know how Barrett will see the tradeoffs. I just know that either way, AMD wins. AMD can either take new market share with mobile or with desktop processors, depending on the degree of Intel's reaction. The time when Intel could subsidize low desktop processor prices with high mobile parts seems to be drawing to an end.
I think we will see the day that AMD's and Intel's net margins and gross margins crossover because of mobile pricing alone -- and I think it will happen by the end of 2002. If that does happen, it will ensure the end of Intel's ability to set prices unilaterally, and the possibility of AMD to have a lot more control over their own pricing.
As long as the 130 nm ramp continues to go as planned, AMD looks like it is in good shape for the next few quarters.
Join the best community on the web! Becoming a full member of the Fool Community is easy, takes just a minute, and is very inexpensive.
AMD's announcement of four new mobile processors gives us a glimpse of what the rest of the year will be like. AMD has said repeated times last year that their goal is 50% or more of the mobile PC market. They also said that all new "net capacity increases" will go to mobile; this will be several million additional processors per quarter -- a significant portion of the 7 million per quarter mobile processor market. If they can continue to broaden their product line (add T&L and SFF) and continue to increase performance, I think roughly 50% mobile market share by year-end is possible, though they may lose desktop share to get there.
Become a Complete Fool