The plain and simple fact is that Advanced Micro Devices (AMD +11.44%), the last player standing in the fight against Intel (INTC +13.96%) for Windows PC processor dominance, has been in a pretty rough spot. The high end of the PC chip market, including corporate buyers, has gone to Intel at this point, leaving AMD with the low end/value chips.
AMD's low-end niche is already at significant risk
One of the principal reasons that AMD has fared so well in the low end is that Intel largely exited that market around 2010. Intel's Atom products, targeted for this space, were not terribly competitive, and AMD's execution in this space was quite good. Intel's share in many of these sub-segments -- addressed principally with cut-down versions of its high-end PC CPUs -- is sub-50% per the General Manager of Intel's PC Client Group, Kirk Skaugen.
Unfortunately, while Intel's high end has managed to hold stable, the mainstream and value segments of the market have faced a number of secular headwinds. The biggest problem, by far, has been the cannibalization by tablets. This affects both Intel and AMD, but the effect is more pronounced for AMD given its much larger exposure (as a percentage of its PC revenues) to the low end versus the mid-range/high end. However, if you thought it was bad for AMD before, it gets worse.
Bay Trail-M already taking share, Braswell could finish the job
I've been calling for share gains at the low end on Intel's part as a result of Bay Trail-M for quite a while, and Pacific Crest's checks seem to confirm this. This platform is significantly cheaper than any of Intel's prior low-end attempts and, at the same time, has a much better cost structure that leverages Intel's tablet efforts. It also appears to offer a significant performance/watt and, quite possibly due to Intel's favorable cost structure, performance/watt/$lead.
Lower power, higher performing, and cheaper to implement. What's not to like about Braswell? Source: Intel.
Things get worse during the back-to-school season and into 2015, as Intel refreshes the Bay Trail-M line for back-to-school, and then launches a next generation part known as Braswell -- which lowers platform bill of materials and dramatically increases performance -- built on its 14-nanometer manufacturing process.
This should give Intel a decisive performance/watt lead over anything AMD will have in the market during that time frame. AMD could try to compete on price, but it's likely that Intel will also be very aggressive on pricing these parts, as well, especially since AMD has to pay an external foundry while Intel gets to "keep" that foundry margin.
Is AMD's PC chip business finished?
At this point, it's hard to see AMD's PC business as remaining long-term viable. As long as it's losing share in a declining market, the financial impact will be quite severe, especially as the company still has significant exposure to PCs. Should the market stabilize/return to growth, then AMD is still faced with a share loss story that will become increasingly difficult to overcome, especially in light of Intel's structural advantages (manufacturing lead/scale, better funded R&D, stronger brand).
Things just don't look good for AMD's PC business and, unless AMD can generate material revenues in non-PC businesses that compete in segments that Intel (and some of the larger ARM based players) aren't already interested in, things just don't look good for AMD as a whole, either.
