The so-called "skinny camel is bigger than Ma Da", Intel, who has been frustrated in the field of mobile processors, has been seeking to turn over, but as Intel abandons the Atom processor product line, let us see that Intel has no choice but to give up in the mobile processor market. . But last year Intel's Altera (FPGA editable logic chip company) revenue pocket showed that Intel is pushing the treasure in the Internet of Things and data center market.
Compared with the mobile terminal market, the size of the Internet of Things market is tens of times its capacity, which is completely determined by the characteristics of the Internet of Things. According to a two-year follow-up survey conducted by BIIntelligence Research, a subsidiary of the US Business Insider,
By 2020, there will be 34 billion devices connected to the Internet worldwide, and in 2015 only 10 billion devices will be connected to the Internet. IoT devices will account for 24 billion units, and the remaining 10 billion devices will be traditional computer devices (such as smartphones, tablets, smart watches, etc.).
At the same time, ARM officially announced on May 19 that it will acquire the British startup Apical for $350 million. Apical is a well-known visual imaging technology processing company. In fact, many mobile phone manufacturers are adopting screen display technology and effectiveness. Apical's technology.
Intel in the PC era and ARM in the mobile era have firmly held the top position in the industry in their respective fields. Although they are all infiltrating each other, the real short-term confrontation eventually appeared in the Internet of Things industry. The first step in solving a problem is often to recognize the root of the problem and to get to the bottom. Where is the core problem of the Internet of Things? From these issues, the author believes that the strategic differences between ARM and Intel can be seen from these issues.
Are the so-called agreements and standards a stumbling block to the development of the Internet of Things?
The Internet of Things, which is known as the third information industry revolution, has been placed in high hopes. In the traditional Internet field, from the PC to the mobile era, the boundaries of the Internet and the Internet of Things are disappearing, and this transition is related to all walks of life, from the consumer field. In the industrial field and even in the agricultural sector, both production methods and lifestyles need to be slowly adapted and adapted.
Of course, specific to the technical problems, the Internet of Things industry now seems to be a mess, from inter-device communication protocols, such as WiFi, Bluetooth, NFC, etc.; to 5G network communication technology standards; then to the operating system platform, etc., there is no Achieve unity. However, the author believes that these so-called standard agreements do not have the relationship of "you die", and they are likely to coexist in the foreseeable future. For example, 5G communication technology will be the same as 4G and 3G eras. There are multiple networks and multiple systems, and the operating system is difficult to unify. The killing of giants such as Apple, Google and Microsoft is difficult to distinguish. These lower-level protocols or standards and API interfaces are fully compatible, and the experience is seamlessly unified at the upper user experience level.
The real flaw in the Internet of Things lies in the shortcomings of the upper-level user experience. The highest realm of the Internet of Things is to help people better manage the whole society. From the city to the family, it needs to face all kinds of problems. How to better? Human thinking and emotions deal with the contradictions of human society, and let the whole society behave in a healthy way is the problem that the Internet of Things needs to solve. Otherwise, some problems with standards and agreements are reversed, and the actual application is still a castle in the air.
At present, the practical application of the Internet of Things requires two problems to be solved, the actual sensing ability, the ability to have visibity, and the ability to transmit data quickly and without delay.
ARM enters the Internet of Things to break through the perception layer
Some time ago, the Alpha dog became famous, so that the technology industry began to re-recognize the artificial intelligence stuck in the bottleneck for decades, but netizens began to tease: If the Alpha dog comes, we can justify the dog. In fact, in addition to helplessness, there is more to be thought-provoking. Artificial intelligence is ultimately "networked", which is the core of the Internet of Things.
Networking is the foundation, but it does not constitute a necessary condition for the Internet of Things. The biggest feature of the Internet of Things is that it can carry out intelligent perception and pre-judgment, and meet the needs of people with the thinking and perspective of "people". Very simple, such as focusing on the scene of smart home, each interconnected intelligent hardware does not even need our deliberate input instructions, but rely on its intelligent sensing functions, such as visibility (visibity), equivalent to human eye function. Accurate perception can guarantee a crucial role in the next step of pre-judgment.
Taking the human sensory function as a metaphor, the five basic perceptual functions of vision, smell, touch, taste and hearing, with the current IoT sensor technology, “visual perception†and “auditory perception†are most likely to be applied on a large scale. Because this represents two mature technologies: camera acquisition and voice acquisition.
And ARM's acquisition of visual computing company Apical is clearly dedicated to solving the practical application of the visual perception layer of the Internet of Things. ARM's IP licensing model determines that it must be able to come up with the most basic architecture in the IoT industry. As a prominent feature of the Internet of Things, the IoT IP architecture with perceptual capabilities will become the standard configuration for ARM licensing.
However, even if the Internet of Things perception function can accurately capture images and sounds, in terms of people, this is only a part of the process. The feedback process of the entire IoT system and humans for a certain action is basically the same. Perceive changes → feedback brain → release response instructions.
How to accurately predict the meaning of different actions of users, and give correct feedback to enable the Internet of Things to complete a cycle, so ARM tries to solve the top-level perception function. As for downstream cloud computing and local computing, even ARM is still difficult. touch.
Intel optimizes IoT local computing
Altera's acquisition of Altera focuses on FPGAs (Field Programmable Gate Arrays). Compared with common ASICs, FPGAs are highly editable and customizable to meet the needs of different users. And real parallel operations (or heterogeneous operations) are implemented without having to perform sequential operations like a PC chip.
In the Internet of Things era, parallel computing can greatly reduce the latency and improve real-time response speed. In addition, FPGAs also have the advantage of low power consumption. As we all know, the Internet of Things does not emphasize computing performance compared to smartphones and computers. Instead, it focuses on power consumption in an extreme way. In the context of the Internet of Everything, real-time low power consumption. Local computing will become mainstream, ensuring "always on". Therefore, FPGA is an ideal choice for IoT local computing.
Of course, the intelligence of the Internet of Things lies in local computing + intelligent cloud computing. That is, local computing performs operations that collect processing data and execute commands, while cloud computing needs to combine artificial intelligence (AI) through deep learning. A smart way to adapt to the needs of humanity.
In addition, because of the multi-channel, fast parallel computing capabilities of FPGAs, FPGAs are also favored in 5G communications, which are known as the Internet of Things.
At the 2016MWC show, Intel has demonstrated 5G communication base stations based on Altera's Arria 10 series FPGA chips. It can be said that Intel's chess game in the IoT industry has basically begun to appear. However, the focus of Intel and ARM is different. ARM's focus is on providing solutions to IP and IP licensing. Intel is still designing and manufacturing all-inclusive.
In comparison, ARM and Intel are still on their own path in business model, relatively open and relatively closed. Compared to ARM, Intel's biggest problem is that it is not good at integrating SOC, and the advantage is limited to the CPU. The Internet of Things will still focus on the integration of the system SOC chip, whether it is WiFi, NFC, Bluetooth and other communication chips or baseband chips, will be highly integrated. So Intel's IoT plan remains to be seen.
The chip giants of the two eras will eventually have a fair duel in the Internet of Things era. Without a first-mover advantage, how will this war finally come to an end? Perhaps the answer given now is only a joke. The author really looks forward to how the Internet of Things will open the third wave of the information industry revolution in the competition among the giants.
Longhua Manxueling Trading Company , https://www.mxlvape.com