The world of PC Vista Unlimited Intel why in the Mobile field Waterloo?

Since 2013, Intel has been working frequently in the mobile field, and has worked with Lenovo Asus to launch its products in the smartphone sector. In the Flat field, but also hit the money below the cost to China's various road shanzhai manufacturers, hoping to get 40 million of shipments. But it backfired, whether smart phones or tablets, Intel is not, why? X86 's problem is that Intel is no stranger to mobile phones, and the Intel PXA series processors have been the best and fastest mobile processors on the market in the early days of the iphone, ARM11 performance in the ARM9 era. Many many years later, ZTE once hit U880 used the core of PXA. Intel was unlucky enough to sell the mobile sector in 2006, and Apple released the iphone in 2007, sparking a mobile internet revolution. And when Intel knows it's time to move on to the mobile Internet, the best time has passed. On PCs, Intel is allied with Microsoft, X86 processors and Windows systems to build a rugged fortress. In the mobile arena, arm and iOS, Android have built up another fortress, Intel's position from defense to attack. Intel wants to move into the mobile domain, first of all to solve the problem of X86 and Android compatibility. For this reason, Intel spent more than two years. And when Intel is finally able to run the vast majority of Android apps without obstacles, ARM has made great strides over the past two years. Intel's sad discovery is that although its new core performance is excellent, the ARM lineup does not have an overwhelming performance advantage. And the performance of smartphones over the past two years has been excessive, even if Intel's chip performance has some advantages, users can not experience. On the other hand, Intel consumes a large number of transistors for X86 compatibility, and does not have an advantage in heating power. To X86 the instruction set, Intel has been more than a few. Ii. fusion of baseband the baseband processor is the core of the smartphone, and the application processor is attached. So, over the years Qualcomm, MTK galloping, TI and Nvidia struggling. Intel previously had no baseband processors, so Intel was unable to enter the smartphone arena, so Intel acquired Infineon and got baseband technology. But sadly, the fusion of communications chips is far more difficult than Intel imagines. Infineon was acquired in 2010, and in the year 2014 there is no SOC for fusing baseband and application processors. Without integrated Soc, it is difficult for the manufacturer to develop, the product quality is stable, and Intel's competitiveness is poor. And in 2014 years, the market has not only a high pass a competitor, MTK has been able to make performance-class chips, Heiss, Samsung also solve their own baseband problems. Intel's slow baseband fusion process has left it lost in the smartphone marketTo be competitive. Iii. the pain of a cheap solution traditionally, Intel is just a chip manufacturer, although in the PC era, Intel also produces well-made motherboards. But Intel has never done cheap cost-control solutions. This has made Intel touch the wall in the tablet market. In the Shanzhai tablet market, while Intel has priced processor prices below cost, without a cheap solution, flat prices are down. Intel later worked with Rui-core to try to solve the problem. Iv. Intel's Choice problem in fact, Intel is not a power issue, but rather a choice. Intel also has powerful CPU design capabilities, with the world's top manufacturing processes. If Intel were to design and manufacture an ARM processor, Apple would not necessarily be an adversary. Should we stick to X86 in the mobile field or give in to the situation using the arm instruction set? In addition, Intel is accustomed to high profits in the PC and server markets, to low-cost competition in the mobile market and lower profit margins. Even if it is possible to pull down the high margins of their PC and server markets, will intel reduce profit margins for the mobile market? To solve the mobile dilemma, Intel has to make a choice first.