Can AI handsets get rid of stagflation market?

category:Internet
 Can AI handsets get rid of stagflation market?


For the smart phone market, the overall screen seems to have become the product of the last century. If we dont take artificial intelligence at the press conference now, Im sorry that the next wave of handset shuffling may soon be killed by other manufacturers. Since the second half of 2017, almost all mobile phone products have been launched with the concept of AI, for example, Samsungs AI assistant, Bixby, can perform intelligent translation and exchange rate conversion through a camera. HUAWEI has also released the first chip to integrate the NPU neural network processing unit, followed by the Mate10, P20, glorified V10, glory 10 and other products that add the features of the photographers, and the glory 10 will be able to learn deeply about the professional photographers views, parameters, and composition experience. Even the low-key apple, also added to the chip bionic gimmicks, said it has a neural network engine with 600 billion times a second, and the use of the word neural network has also led to a wave of mobile phone conference PPT this year. But AI is more like a marketing gimmick at the moment. The same technical rendering does not bring more good news to the market in the first quarter. According to the March 2018 domestic mobile phone market operation analysis report issued by China Institute of information and communications, the shipments of domestic brand mobile phones in the first quarter fell by 27.9% in the first quarter. AI is not a function, a module, or a technology, but a complete system composed of chip algorithms, systems, and so on. If a company wants to build a real AI capability system, the three levels of hardware, intelligent systems and applications are essential. Zhao Ming, the president of glory, pointed out at the GMIC global mobile Internet Conference that there are a large number of follower AI in the current industry, that is, the packaging AI with no essential upgrading, the hype AI with no substantial landing, and the lack of closed AI for ecological construction. The geometry of gold content in new products Huang Yuxuan, a research manager of the state advisory action memory, told the first financial reporter on 2 days of the first financial reporter that in the field of intelligent mobile phones, the upstream manufacturers in recent years include AI applications in the chip development, believing that the application of AI will be more popular, and the antennae will also extend down to the medium - end mobile phone. But if the brand wants to travel alienation, it must invest huge funds to build AI applications. In his view, AI is divided into three levels, including the first layer of chips and hardware, for the core processing of AI to build the hardware suitable for the AI algorithm, and the second layer to build the AI intelligent system, enhance self-learning cognition, computer vision, and third layers, application ecology. At any level, the bonus of this industry can be enjoyed. But in fact, more AI phones in the market are still in the follow suit marketing status. For example, using the older chip architecture, without the essential upgrade, but packaging the third generation, dual core and other AI chips concept, the actual computing force is far behind the industry level. For example, in the shooting function of the name of AI, the publicity is greater than the actual experience, but the actual picture effect is not good, and even some only in the conference PPT packaging scene identification classification, in the mobile phone function of no embodiment. Indeed, many AI phones emphasize the ability to focus on image recognition and processing. Huang Yuxuan told reporters, for example, to open a camera, if you aim at one person, then a small icon of man will appear in the lower left corner of the preview picture; if you aim at the plant, there will be a small icon of a plant. More specifically, the AI is still a baby state, which requires the user to send instructions, the AI on the phone to play a role, and the passive response. At present, the development of AI is mostly based on single AI applications, such as speech recognition, smart environment sensing and taking pictures. Huang Yuxuan told reporters. But in fact, AI will become a coprocessor of human brain in the future and shorten the gap with experts. Zhao Ming told reporters that AI technology can make mobile phones study themselves based on user behavior habits, and then optimize, intelligently perceive user scenes, predict user behavior and allocate resources intelligently. More intuitively, with the application of AI in the mobile phone, the mobile phone starts to observe you and understand you, the phone is just a tool, but in the future it will become a brain, no longer just a hardware state. Can save the stagflation Market With the intensive release of AI mobile phone products, a AI mobile marketing war came roaring. However, whether it is the public statement of the mobile phone manufacturer, or the data revealed by the research agency, the anxiety mentality and the market posture of the glue are not relieved in the first quarter. On the contrary, more and more small and medium-sized mobile phone manufacturers have become more difficult to survive. And the head enterprises defensive attack makes more and more small and medium-sized brands face the pressure of declining share. From the data rankings provided by the analysis agency, we can see that the Chinese market has a distinct polarization in the first quarter, and the top six manufacturers are all shipments in the first quarter of 10 million, starting with the seventh Meizu, but the shipments of Jinli, Samsung and chili are all million. This also means that the mobile phone manufacturers who rank outside ten have less than one million shipments in the first quarter, with an average sales volume of less than 400 thousand per month. For small and medium-sized brands, the severe winter market is an indisputable fact. In the view of many analysts, it is unclear whether AI can save the stagflated mobile phone industry, but it is undeniable that the combination of mobile phones and AI will be one of the future technology vents. Yu Chengdong, the head of the HUAWEI terminal, once called the chip of the integrated NPU artificial intelligence processing unit as the nuclear weapon in the chip, and a mobile phone chip behind thousands of peoples R & D team, both master and doctor, integrated the NPU of the dedicated independent hardware processing unit, and completed the special AI operation task independently by NPU. To improve the AI computing power. In contrast, high pass uses three modules, CPU, GPU and DSP, to perform AI operations. Through the neural processing engine (SNPE), the specific AI tasks are analyzed and sent down, while the Federation uses a dual core APU to improve the operation efficiency of the image post-processing through the ability of multiple DSP, and then do a AI related image processing algorithm. Like AI, chips are all heavily invested, and one shot is ten years, ten years to build the chip that is leading today in AI. Zhao Ming told reporters that in the full comparison and choice, the stronger will be stronger, and the concentration of mobile phone brands will be higher and higher. Source: first financial daily editor: Yao Liwei _NT6056 Yu Chengdong, the head of the HUAWEI terminal, once called the chip of the integrated NPU artificial intelligence processing unit as the nuclear weapon in the chip, and a mobile phone chip behind thousands of peoples R & D team, both master and doctor, integrated the NPU of the dedicated independent hardware processing unit, and completed the special AI operation task independently by NPU. To improve the AI computing power. In contrast, high pass uses three modules, CPU, GPU and DSP, to perform AI operations. Through the neural processing engine (SNPE), the specific AI tasks are analyzed and sent down, while the Federation uses a dual core APU to improve the operation efficiency of the image post-processing through the ability of multiple DSP, and then do a AI related image processing algorithm. Like AI, chips are all heavily invested, and one shot is ten years, ten years to build the chip that is leading today in AI. Zhao Ming told reporters that in the full comparison and choice, the stronger will be stronger, and the concentration of mobile phone brands will be higher and higher.