When Advanced Micro Devices (AMD 3.05%) touted the launch of its low-power Kabini and Temash processors, many believed that these would finally be the products that helped turn the tide in the rather unsuccessful war against Intel (INTC 3.03%) in the PC space. While these parts weren't necessarily aimed at mounting an assault on Intel's high-end PC beachhead, the former was supposed to attack the fairly large mid-range PC market and the latter was supposed to gain the company a foothold in Windows 8.1 tablets. Contrary to positive fanfare at the launches of these processors earlier in the year, it seems that these products will ultimately do very little to change the competitive dynamics in either PCs or tablets.
Some background
AMD, like Intel, develops two lines of processor cores – a "big" core meant for higher-end notebooks, desktops, and servers, and a "small" core meant for low-cost PCs, tablets, and game consoles. AMD's "big cores" have continued to lose share across key market segments as the company's designs simply can't compare with what behemoth Intel is able to leverage. It's very tough to compete by cutting prices when your competitor can build faster, less power-hungry, and cheaper-to-make designs than you can.
On the small core front, AMD has actually done well. Years ago, AMD's "Bobcat" processor was a superior design to Intel's low-power Atom. A major part of this was likely due to Intel's unwillingness to risk cannibalizing its own "big core" sales, but another non-trivial part of this was that Intel did not leverage its manufacturing lead on the "small" cores.
Today's Kabini and Temash system-on-chip designs are based on the successor to the Bobcat, known as Jaguar. While chips based on these designs seem poised to offer excellent performance/watt and performance/dollar, there are a couple of things that cannot be ignored here.
PC share gains unlikely
AMD's Kabini is a competent processor – it competes well with the very lowest end of Intel's processor stack, and offers competitive performance per watt. Unfortunately, while Intel has traditionally serviced the low end of the notebook market with cut-down versions of its larger Core processors, which are more power-hungry and more costly (due to being a two-chip solution against the single-chip Kabini), Intel is ramping up its next generation Bay Trail-M processor.
This chip is built on Intel's 22-nanometer process with FinFETs (that means lower power and a smaller chip than what AMD can do on TSMC's 28-nanometer planar process) and generally offers a performance and power advantage over its AMD brethren. While it had the low-cost/low-power PC space mostly to itself many years ago, Intel's push into the tablet space has also given it a weapon with which to aggressively fight in the lower end of the PC space. While Intel had yet to roll out these processors in early Q3 (yet AMD had for notebooks, which means that AMD was able to gain some share in the low end), the fear is that with Intel ramping up its Bay Trail-M products at the end of Q3 for sale in devices during Q4, that share gain will reverse and continue to worsen.
The tablet situation looks worse
While AMD could theoretically stave off share loss with extremely aggressive pricing in the low-cost PC space since its products are more or less substitutes for Intel's, the situation is different in the tablet space. AMD has been talking up a ~4 watt dual core Temash part, and there are fundamental barriers that belie this seemingly appropriate headline power consumption:
- AMD's Temash does not include an integrated Image Signal Processor, which is an IP block that controls and provides functionality for the built-in camera.
- Temash for tablets includes just two Jaguar processor cores running at 1GHz. Per core/per clock, Jaguar has some advantages over the Silvermont core in Bay Trail-T, yet the difference isn't close enough to make up for half the cores running at 42% of the frequency. Silvermont also has more aggressive power management.
- In 3D Mark Ice Storm, a 3D graphics benchmark, AMD's tablet-oriented Temash scores 12,495, while Bay Trail-T has shown scores in excess of 13,000. This means that AMD doesn't even have a graphics advantage.
- No tablet design wins have been announced in the 10" and smaller form factors.
In short, AMD is – as it is in the PC space – still not competitive enough with Intel to make meaningful inroads in the new era of mobility.
Conclusion
AMD's strategy to diversify outside of the traditional PC space cannot happen fast enough. Intel is far too potent a competitor with too much on the line to make life easy. That being said, AMD is the chip supplier for Microsoft's Xbox One and Sony's Playstation 4, so while this business isn't high margin, a successful console ramp up could help offset AMD's traditional PC woes. And, who knows, maybe the partnership with ARM to leverage its IP in future server chips could work out. Only time will tell, so AMD remains a high-risk speculative bet.
