On the other hand, Huawei has just launched the world's first mobile phone equipped with an AI chip, the Unicorn 970, capable of "deep learning." On the other hand, Qualcomm, which ended the 4G/5G summit, worked non-stop with AI vendors to optimize the AI ​​algorithm from the chip layer.
If the three-generation AlphaGoZero that recently blew up the circle of friends just made the masses "unclear," then the AI ​​handset seems to be getting closer and closer to us. In fact, mobile phones using AI have become commonplace, such as filters for live broadcast software, face recognition for payment software, or voice assistants such as Siri and Xiao Bing. But if the software can already solve the problem of mobile phone AI, then there is still market space for the AI ​​chip? ''
AI mobile phone calculations
Problems with battery life
“Actually, the user’s perception is not obvious, but AI has been improving. For example, the beauty must be shot before you can see the photo, but it can be previewed in real time,†said Xu Li, the CEO of Shang Tang Technology, who told the Southern Reporter that AI has been making progress. These changes require the acceleration of a similar high-pass chip DSP.
"But, for example, the time-delay problem of face unlocking; for example, it is now possible to face-face, but still can not do thin thighs, such as taking a lot of people's photos, can not do for each person's face tagging recognition. According to Xu Li, the impact of AI on the terminal is still relatively weak, mainly because of the lack of computing power. To solve the AI's processing, UHF calculations are needed, otherwise the delay is very serious; in turn, the amount of calculation is too large and face power consumption problems, which is the problem of cell phone heat and life. From the data on the Qualcomm 4G/5G Summit, it is easy to see that 89% of the users are now the biggest pain point is the battery life.
Some vendors choose to put computing power in the cloud, such as Intel. "The carrying capacity of the terminal hardware will become the major bottleneck of future computing power. The demand of AI will make the phone become a 'big stove.'" Wu Geng, chief technical expert of the wireless standard of Intel Academician and Intel Communications and Equipment Division, told Southern Reporter that the future There is no so-called high and low allocation of mobile phones, and the effect of a large memory bar is not significant.
However, Xu Li believes that most of the calculations also need to be completed locally. "If many people use AI programs at the same time, they need a lot of bandwidth for cloud transmission; and delays will be more serious." And Meng Tong, chairman of Qualcomm added, "For example, privacy, users do not want to complete in the cloud. â€
Power consumption: AI chip
Should it be a plug-in or an integration?
If local computing is important, how can we avoid "fever"?
Terminal manufacturers choose to plug in a new module. This also requires a new understanding of the chip architecture. The traditional mobile phone chip mainly consists of three parts: CPU (Central Processing Unit)/GPU (Graphics Processor)/DSP (Digital Signal Processing). The Huawei mate10's AI chip is based on the traditional core architecture, adding an NPU (neuron module) developed in the Cambrian era to reduce the CPU/GPU burden and reduce power consumption. Coincidentally, the vivoX9sPlus released in July this year is also equipped with an independent DSP image optimization chip to improve the camera technology.
"If it is an entry-level low-end mobile phone, it can solve massive data through an independent chip," Keithkressin, senior vice president of product management at Qualcomm, told Nandu reporters, but before that, we must also consider AI processing capabilities, independent chip prices, software models, and power management. A series of questions. "The company's flagship machine will still choose to integrate with Qualcomm, because the independent chip will increase the size and cost of mobile phones."
"Compared with custom hardware, flexible artificial intelligence hardware is currently a better choice, combining DSP and GPU is better. We spend most of our time on the SDK and framework." Asked to be the first to release AI chips by Huawei. In his opinion, Keithkressin said that it is not difficult to put AI's cell module on the chip, and Qualcomm can do it before. "The key is still in the software development, programming module optimization."
Xu Li also believes that the independent A chip needs a longer development time. “It is not necessary to have a dedicated AI chip. The key is to release more computing resources and other modules to handle complex scenarios.†The cooperation between Shangtang and Qualcomm is mainly in areas such as innovative vision and camera-based graphics processing. "The integration of hardware and software development is the trend of AI mobile phones. The key is to break the middle development layer (SDK) barrier between the low-end and end-applications of the chip," Xu Li told Nandu. "The previous chip layer mainly optimized the operating system, but Now AI requires the chip to directly optimize the application."
A business tablet must have powerful business functions. At present, users use tablet products more for entertainment, and insufficient development of commercial value has become a major factor restricting the development of tablet computers. Many far-sighted terminal manufacturers have set their sights on the business tablet field.
Generally speaking, when business users choose tablet PC products, they still focus on "regular items" such as processor, battery, operating system, and built-in applications, especially office software applications, which are a good helper for business users.
First of all, there must be powerful office software such as customer management software, which can record customer data and related information anytime, anywhere. There is also a need for powerful marketing software, software marketing, reducing manual intervention, and one person can handle all the publicity and customer information.
Business Tablet,4k,5G
Jingjiang Gisen Technology Co.,Ltd , https://www.jsgisentec.com