Ideepwise Wants To Build Artificial Intelligence Machine Brain

As technology matures, people's lifestyles are becoming more intelligent. According to Research and Markets, the global market for artificial intelligence is expected to grow by more than $ 500 billion by 2020. But the current AI and Robot has two common problems, one is the local terminal without "artificial intelligence brain", relying on cloud server computing; the other is based on the depth of learning algorithm model in the terminal computing performance is not high, heavily dependent on the line Server performance, real-time poor performance. To this end, deep thinking artificial intelligence also gives its own solution.

Deep thinking of artificial intelligence, was established in August 2015, is a focus on artificial intelligence core algorithms and special chip technology companies. At present the main product is ideepwise AI Robot Service bottom service system, based on the bottom of the service system IDeepwise artificial intelligence "medical brain" and second generation artificial intelligence robot system and FPGA AI acceleration chip. November 2015, deep thinking had won 1.5 million angel investment Xin, and then in February this year, nine shares of 10 million Pre-A investment.

Mentioned the field of entrepreneurship in the field of AI, the most important thing is how to solve the technology from the product to the commercialization of the problem, deep thinking is no exception. Founder Yang Zhiming said, with many AI start-up companies, deep thinking in the past this time also do a lot of exploration, business model from the original 2C mode to 2B direction, from the beginning of the multi-scene coverage to focus on Specific vertical field. In this process, deep thinking also gradually find a place where they are best at the best and the most suitable landing scene. At present, the company mainly for medical, automotive and intelligent home three areas to provide artificial intelligence products and services solutions.

(NLP / NLU) based on the depth of learning to achieve the vertical field of multi-round multi-task semantic interaction and Chinese proofreading, based on the machine vision (based on the depth of the study of artificial language, Multi-modal emotional computing with text semantics and FPGA-based depth learning performance optimization and acceleration. About the chips, you will know XC7K325T-1FFG900I and JS28F128J3D75.

At present, in the medical field, deep thinking is based on the provision of IDeepwise artificial intelligence based on the "medical brain" of the two products, one is Ideepwise artificial intelligence assisted interrogation system, that is, the use of artificial intelligence to achieve multiple patient consultation consultation , And according to the patient's question automatically given the content and recommendations, the doctor confirmed that the artificial intelligence given the contents of the reply and recommendations to respond to patients, greatly improving the hospital and doctors interrogation efficiency; the second is TCT cervical cancer artificial intelligence auxiliary reading Tablet system to help read the doctor to achieve the TCT cervical cancer screening of the film, the current sensitivity and specificity has been infinitely close to 95%, more than 65% of human reading sensitivity and 92% specificity. Compared with the new Bo's reading system only gives 22 field of vision, deep thinking of the reading system can be given directly to the TBS final grading report, more to reduce the workload of reading doctors and reading efficiency. In the field of automotive, through the cooperation with the OEM manufacturers will be combined with the vehicle to achieve the control, to provide second-generation artificial intelligence robot system, the system can achieve driving language natural language human interaction, traffic reminders, traffic conditions and fault reminders and asked Diagnosis, according to the driver's customary personalized service. In the field of smart home, through the cooperation with the smart home manufacturers to provide based on ideepwise second generation of artificial intelligence robot "thought" service products, and Service authorization and smart home products, through the use of artificial intelligence instead of manual, fault diagnosis , Natural language control and other functions to simplify the operation of household equipment.

And it was able to do this, Yang Zhiming also told 36 krypton, deep thinking not only in the key technology to do in-depth research and development, in the accumulation of data also do the depth of accumulation. The domain of the ontology knowledge base capacity already contains music, medical, automotive, smart home and health recipes, including nearly 100 million ontology knowledge, Chinese proofreading, semantic identification has also accumulated 5 million users.

In the marketing, deep thinking at this stage in the automotive, smart home, medical products and services by the deep thinking of marketing business team to develop their own cooperation with agents to promote the way. From the profit model point of view, the current main provider of robot brain artificial intelligence software saas services and software and AI acceleration chip all in one integrated solution services. It is reported that the company has achieved close to 10 million revenue.

If you want to learn more informtaions about chips,please go to electronic component distributor.

Intel PK Qualcomm,Who is The Internet Chip Overlord

If Intel dominated the PC era, Qualcomm is the king of the mobile era, which one will become the Internet chip overlord on Internet of Things?

Billions of dollars, almost all authoritative market analysis agencies, chips and communications manufacturers, the Internet of things made a positive market forecast in the next few years.

From the market research firm MachinaResearch point of view, the number of Internet connectivity and Internet of Things revenue between 2015 and 2025 will triple, increased to 3 trillion US dollars, and by 2025 the number of global Internet of things will reach a staggering 27 billion.

The forecast from IDC is even more optimistic that by 2020, the number of Internet connectivity will reach 28.1 billion, the size of the income will be more than 7 trillion US dollars. Communications equipment manufacturers Ericsson believes that before 2021, the number of global Internet of things equipment connections will be 23% compound growth rate continued to rise, by 2021 will be 15.7 billion Internet of things connected.

The most daring forecast from Intel, Intel Outlook By 2020, there will be more than 50 billion machines and equipment interconnected, in addition, there are more than 200 billion networked sensors, human society will find the resulting data with our current experience to the data is not an order of magnitude. For example, a networked car can generate 4 TB of data per day, and an Internet of Things plant can create 1PB data per day. How much 1PB actually represent? Assuming an MP3 stored in the 1PB size of the ordinary format songs, then the continuous play to listen to these songs need to spend more than 2000 years. There are also many electronic components distributoron the market.

Things networking technology covers a wide range of complex scenes, there is no way only through a technology, a network, a system to provide all the services, all things are very diverse, the technology is also very flexible, You need to configure the required communication and computing power according to the actual situation. "High-pass product marketing director Shen Lei to the" IT Times "reporter explained, such as Bluetooth, NFC, ZigBee and other short-distance connection technology; and Wi-Fi technology itself has various grades, including 802.11ac, 802.11a / b / g / N, 802.11ax, etc .; there are WAN connection, the wisdom of the city can take all the lights and parking are connected to the Internet, which requires cellular technology, which is familiar with the 2G, 3G, 4G and the future 5G. There are proprietary, non-authoritative technologies specifically for the networking of technologies such as Sigfox and LoRa.

As blossom of the technical pattern, so that things the standard dispute has become increasingly confusing. At present, the global authority of the standardization organization 3GPP latest release version basically established three kinds of things on the wireless connection technology: NB-IoT (narrowband Internet of things), eMTC and EC-GSM. These three standards, it is precisely to replace the unauthorized low-power wide area network technology (such as Ingenu, LoRa and SIGFOX), which makes the latter's industrial chain status crumbling.

You may interested on: LM2576SX-5.0 and K9F4G08U0D-SCB0