Nearly 20 billion dollars in 4 years! 9 investment and 6 M & A, Intel enters the medical market strongly

Not long ago, such a news attracted our attention - on August 5, 2019, Intel Capital, an investment institution of Intel, led the b-round financing of exo imaging, a medical imaging company, with a financing amount of US $35 million After the news, Intel's share price did not fluctuate much on that day, but the trading volume of that day reached a new high in the past three months Before long, on August 19, mecamand robot announced that it had obtained Intel's investment The start-up enterprise had deployed the intelligent solution of "3D + AI + industrial robot" in the hospital pharmacy Before that In April 2019, Intel investment led the round a financing of two medical start-ups, namely, reveal Biosciences, a computer pathology company, and medical information, which provides remote medical data monitoring and analysis software For the first time, the layout of the two companies has made the Intel investment global summit, which once reported investment results to investors, separate the investment directory of medical treatment What's more, Huiyi Huiying, a domestic medical AI head enterprise, was invited to attend the 19th Intel Global Investment Summit as a strategic investment partner of Intel and a Chinese medical image AI head enterprise All this seems to show us that Intel, as an IT giant, is quietly starting to lay out in the medical field As a result, artery network (wechat: vcpeat) has a strong interest in Intel's medical layout in the past, and has looked up Intel's news about the medical field in the past 20 years, combing out Intel's medical panorama To sell chips! As one of the largest chip manufacturers in the world, Intel is the main founder of x86 architecture Most desktop and notebook processors, as well as server chips in the computer room, are based on this architecture So, less strictly, Intel has long been associated with healthcare - although most of the time it's passive as a device provider According to well-documented information, Intel's first attempt to actively associate itself with the medical industry occurred on February 24, 1998 One of the earliest press releases available on Intel's official website was a press release: "Intel and communihealth announced a plan to help patients and doctors better manage chronic diseases." Communihealth (now nowhere to be verified) was a company that specialized in providing personalized online health information at that time By partnering with Intel to provide Internet services, diabetics can use PCs to track blood sugar levels, diet and activity, and share this information with commonhealth's healthcare professionals, who then provide medical advice online Considering that in January 1998, Pentium II processor was just released; at that time, Internet service was still in the era of dial-up Internet through telephone line, and the fastest Internet speed was only 52.6kb/s (today's optical fiber network speed can reach 100MB / s, and 1MB = 1024KB), which was a quite advanced move Andrew s Grove (Andy Grove, one of Intel's founders), then chairman of Intel, was keenly aware of the addition of the Internet to medical informatization In October of that year, Intel, together with the American Medical Association and the American Academy of Pediatrics, jointly held the Internet health day Andrew s Grove made an opening speech, predicting Internet driven change in healthcare At that time, both Intel itself and Intel's investment business had been trying to use the Internet as a starting point to get married with healthcare However, it is clear that the famous "Internet bubble" handed over to the century has completely changed this direction Since 2003, the gradual maturity of WiFi technology has made Intel try to change medical treatment with technology again Eric dishman, then Intel's prospective health research project manager, tried to explore Intel's most promising field in the medical field at the future of Aging Services Conference held in that year, focusing on the added role of home sensor network for medical health After a period of exploration, Intel has set the direction in medical treatment On January 17, 2005, New York time, Intel announced a drastic restructuring of the company's structure, adding a new digital health group (later renamed health and life science group, i.e the Department of medical health and Life Sciences), led by Louis, vice president Burns director, whose main task is to explore new applications of Intel processors in medical research and personal health care From this goal of the digital medical department, it can be seen clearly that Intel's original intention of establishing this department is to expand the application scope of its chips After all, this is Intel's core competitiveness Driven by this concept, Intel's digital medical department has developed the MCA platform (Mobile Clinical Assistant) in combination with Intel's chips The motion C5 medical tablet based on this platform was also a very advanced product at that time - until 2010, apple just released the iPad we know well July 10, 2008 became another milestone in Intel's medical career The Intel Health Guide tablet, which integrates a series of medical management software, has passed the FDA 510 (k) certification, and is also the first Intel medical device to pass the FDA certification During the period of Intel Health Guide tablet, Intel also participated in the medical industry through cooperation For example, alliance with GE Healthcare and set up a new healthcare joint venture, care innovations (the company has developed many medical solutions, such as remote patient management and family self diagnosis), funding medical education and medical research projects of young talents, etc All these have greatly enhanced Intel's sense of existence in the medical field From 2013, with the development of technology, Intel began to participate in the medical field in a broader way Intel is involved in telemedicine, wearable devices, IOT, cloud computing, big data and AI In 2012, Google built an AI called "Google brain" with 16000 Intel multi-core processors, and identified cats in YouTube videos, which became a landmark event in the development of AI Subsequently, MIT and broad Research Institute of Harvard University improved the speed of gene information analysis and monitoring by 3-5 times using the AVX vector instruction set of Intel processor Taking this opportunity, Intel found the huge potential of AI in the medical field and the huge advantages of its products in AI, and gradually began to focus the attention of the medical field on AI For example, Intel cooperates with Novartis to use AI to accelerate drug development; with Philips to speed up in-depth learning of medical AI; and with Siemens to explore the potential of AI real-time diagnosis MRI Eric dishman, the soul of Intel medical, is the soul of Intel medical layout He joined Intel in 1999 and has been general manager of Intel's health and Life Sciences Department since 2005 He is responsible for the development of cross departmental strategies, products, R & D and policy projects related to Intel's health and life sciences By the time he left Intel, he was an Intel academician and vice president of health and life sciences His department focuses on improving Intel's growth in medical information, genetic information and personalized medicine, consumer health and collaborative medical technology Eric dishman's experience is legendary In his teens, he was diagnosed with a rare form of kidney cancer He struggled with the disease for 23 years before the successful kidney transplant in 2012 Before that, almost every two years will be informed by doctors that there is still one or one-and-a-half years left It was not until he had sequenced the DNA of the tumor that he found a treatment to help save his life Despite his serious illness, Eric dishman did a good job in his job Disease made him better able to understand the role of technology in medical progress Under his leadership, Intel has built up a medical layout step by step from scratch It can be said that without Eric dishman, there would be no Intel medical layout today In April 2016, Eric dishman was nominated as the head of precision medicine Initiative (PMI) of NIH After that, Eric dishman is now mainly responsible for the all of us research project evolved from PMI The project, involving 1 million people, aims to further study precision medicine The competition between Intel and NVIDIA in medical AI is becoming increasingly fierce When it comes to AI, we can't help mentioning NVIDIA, Intel's inevitable old rival NVIDIA's main product GPU was originally born for 3D graphics rendering In the PC architecture, it is related to Intel's main product CPU, and there is no conflict between them But in AI applications, especially in the medical AI applications where computer vision and voice processing occupy the absolute mainstream, GPU has inherent advantages In 2013, NVIDIA cooperated with Stanford University to build the AI neural network with the strongest performance at that time by using GPU By using 16 GPU servers, the AI performance was increased to 6.5 times of that of Google brain - the 16000 CPUs used by the latter needed to be accommodated by up to 1000 servers, so the advantage of GPU in AI can be seen Up to now, most of the medical AI enterprises have adopted NVIDIA's GPU based solution The head enterprises of domestic medical AI, such as conjecture (2017), Shangtang (2017) and Yitu (2018), have all appeared at NVIDIA GTC conference However, GPU is not always dominant in AI applications Due to the limited memory processing capacity of GPU, CPU is superior to GPU in some occasions where it needs large capacity memory (tens of GB or even hundreds of GB) GPU has no advantage over ASIC designed for a specific purpose, such as AI chip with AI algorithm It is for this reason that in 2018, Yitu technology, which also participated in the NVIDIA GTC conference, released the quest core in 2019, which is specially optimized for the AI algorithm of Yitu, greatly improving the computing power of specific AI scenes It is said that the computing power provided by a server based on a quad core chip is equivalent to that of a server with eight built-in P4 GPUs The size of the server is only half of the latter and the power consumption is less than 20%, which fully demonstrates the huge potential of a dedicated ASIC Intel also saw this In August 2015, it acquired nervana system, a company with a high reputation in the field of AI dedicated chips, and in 2016, it acquired movidius, a high-performance visual processing chip manufacturer Later, Intel released its first AI chip, Lake crest, which claims that AI performance is 10 times better than GPU; the latest generation of spring crest (nervana nnp-l) is said to have increased AI computing power by 3-4 times compared with the first generation of Lake crest At the same time, Intel also developed the Spring Hill AI processor (nervana) by replacing the internal core of core display with the AI reasoning accelerator based on the ice lake architecture adopted by the desktop core processor, and adding a DSP engine to it Nnp-i) - this chip has been released on August 21, and the AI processor that can be deployed seamlessly in the existing data center can support all the mainstream deep learning frameworks AI chip is good, but it costs a lot of money to develop and produce FPGA with special algorithm is more flexible and configurable