Earlier this week, I blogged about the pressures facing Intel as it tries to adapt to a world in which its strength — processors for desktop PCs and other big computing devices — is de-emphasized. Instead, mobile computing, which puts different demands on processors, is well on the road to domination.
An issue that didn’t come up in that piece but will have significant impact on the future of mobile processing is the number of cores that will become standard. MIT Technology Review reported on Dec. 12 that MediaTek, a Taiwanese company, has introduced the MT6589.
The new processor is a quad-core system on a chip (SoC). A SoC is a single element that combines many of the tasks that formerly were handled by discrete processor elements. As the name implies, a multicore processor (dual, quad or beyond) runs processors in parallel in order to perform a particular task more quickly. A quad-core SoC, then, theoretically will create efficiencies on two levels.
The story suggests that the immediate impact will be felt in China. However, nothing in the new world of telecommunications stays national or even regional for very long. MediaTek has relationships with companies such as Huawei, ZTE and Lenovo, which may find the MT6589 interesting. The story says the ramifications could influence overall conditions in the industry:
This, in turn, could force the industry at a global level to undergo another cycle of price competition, which would have a big impact on high-end brands like Samsung, according to IDC analyst Teck-Zhung Wong.
It is important to recognize the importance of emerging markets on the evolution of processors. The subtext of the MediaTek story — that the ramifications of the company’s quad SoC eventually will be felt in other markets — is echoed in a Forbes story on Qualcomm’s introduction of the entry-level Snapdragon MSM8226 and MSM8626 quad chipsets to China and other emerging markets. The piece suggests that emerging markets are the drivers of much of what is happening in chip development.
There is a bit of fuzziness, at least for an outsider, on the relationship between the number of cores and how devices perform. This post at Tom’s Hardware relates what apparently was a panel discussion of some sort between IHS analyst Francis Sideco and Matt Wuebbling, the director of marketing for Nvidia’s Tegra processors. The upshot is that the limit on cores likely will be four and the focus will shift more explicitly to driving the customer experience by doing such things as prolonging battery life.
That piece — and a similar one at CNET — doesn’t fully connect the dots between the number of cores and end-user performance. The CNET piece, which was written by Roger Cheng, suggests that the biggest impact of core saturation will be on the self portraits chip makers, handset vendors and carriers paint. Companies will have to appeal to customers by doing more than focusing on winning the core race:
In its place, expect chip companies, handset manufacturers, and wireless carriers to shift their marketing away from an emphasis cores and more toward tangible benefits such as a smoother experience, bigger displays with higher-resolution graphics, and better battery life.
There are two problems. The first is that Cheng doesn’t clearly say how the number of cores impacts customer experience: Where are more cores helpful, and where do they hurt? The other point is that the vast majority of smartphone shoppers don’t really know what cores are and what they do. I am not aware that the main marketing thrust — at least on the consumer-facing side — has been tied to the number of cores a particular device has.
The mobilization of telecommunications, in parallel with the unleashing of non-developed markets, is having a profound impact on chipsets. Corporate planners should be aware of these transitions, since they will impact — overtly and subtlely — the capabilities of the mobile devices that employees use.