Although shares of Advanced Micro Devices (AMD +0.19%) rose in 2013, they did so only after the stock lost 75% of its value in the previous year. The troubled company has seen its revenue and profit decline. A shrinking PC market and continued share losses to Intel (INTC +2.50%) in the server market have taken a serious toll, and its graphics business is in a perpetual state of being one step behind NVIDIA (NVDA +1.13%). While revenue from the game consoles will help AMD's top line in 2014, some serious problems need to be fixed.
Game consoles may lead to greater things
One bright spot for AMD in 2013 was its sweep of the major game consoles. Both the Xbox One and the PlayStation 4 are powered by custom AMD parts, and it's been estimated that the company receives around $100 per console sold. If the current generation matches the previous generation in terms of sales, a significant part of AMD's total revenue could come from these game consoles in the coming years.
These types of contracts, where AMD designs semi-custom chips for specific customers, are part of AMD's long-term strategy, and it should lessen the impact of the declining PC market. The upside to the game console wins is that revenue is essentially guaranteed for many years, assuming the PS4 and Xbox One continue to sell well. The downside is that margins tend to be low, so it will take a significant amount of additional revenue to make up for falling PC chip sales.
Given the resource gap between AMD and Intel, these semi-custom deals look like the best way for AMD to attempt to return to consistent profitability. AMD's biggest challenge in 2014 will be parlaying the console design wins into other deals and proving that its semi-custom business is not just a one-hit wonder.
Spread too thin
While these semi-custom deals are a good idea, AMD has a lack of focus that will simply not work for a resource-constrained company. AMD sells desktop chips, is trying to push its mobile chips with new launches next year, sells x86 server chips while planning the release of ARM-based server chips in 2014, builds semi-custom chips for game consoles, and sells graphics cards through ATI. The company is trying to compete with Intel and NVIDIA on all fronts, a strategy that is destined to fail.
AMD has spent about $1.2 billion in the TTM period on research and development, a number which has been falling for the past five years. Intel, on the other hand, has doubled its research and development spending since 2008 to $10.4 billion per year, nearly twice AMD's annual revenue. Couple that with Intel's huge manufacturing advantage -- with AMD still utilizing a 28nm process as Intel is pushing toward 14nm next year -- and there's simply no way that AMD can catch its rival. In the desktop and high-end server markets, Intel is untouchable. And in the mobile market, where energy efficiency is extremely important, Intel's manufacturing edge should keep it one step ahead of AMD.
AMD does have the opportunity to grab a bigger share of the microserver market with its upcoming ARM-based server chips; applications that work best with a large number of low-power cores are good match for the ARM architecture. However, Intel offers low-power Atom server chips aimed at this purpose, with Intel's manufacturing edge giving the company a huge advantage. It won't be easy getting companies to opt for ARM-based chips.
On the graphics side of things, AMD competes with NVIDIA, a company focused solely on graphics. While AMD's most recent graphics cards have been a success, causing NVIDIA to lower its prices in order to compete, the fact that NVIDIA was able to have such elevated prices for so long is a sign that AMD is having trouble keeping up. Early in 2014 NVIDIA will launch a refresh of its line of GPUs, built on a new architecture called Maxwell, which promises to be 3-4 times more powerful per watt than its predecessor. AMD will likely have no answer to this for quite some time.
The bottom line
AMD's strategy of competing in all markets simply isn't working, and the company needs to focus on fewer areas instead of trying to best both Intel and NVIDIA on multiple fronts. Intel has an edge in manufacturing and research that at this point is insurmountable, and NVIDIA's sole focus on graphics keeps the company one step ahead of AMD. The semi-custom deals are a good development, but AMD hasn't yet proven that deals beyond the game consoles can create meaningful revenue. There's still hope for AMD, but the company needs to use 2014 to refocus its efforts.
