How hard to rebuild "Jarvis"? Sugon HPC could help

2019-03-06 17:47:54

In Marvel movies series such as "The Avengers" and "Iron Man", it is believed that everyone will be amazed by all sorts of cool scenes. And as we know, Jarvis played an extremely important role in helping Iron Man to win the fighting.

Who is Jarvis? It is a super intelligent software in the film, and also an intelligent housekeeper with ability of self-studying and independent thinking. It could not only help Tony Stark to manage all kinds of affairs and deal with calculation of huge amounts of information, but also automatic accurate calculate the damage rate of the equipment and the accurate time needed for repair when battling or injured.

 

Many people may consider how to "re-create Jarvis" in real life? One of the technologies behind this is deep learning, and that's what we're going to talk about today.

In general, the deep learning is compared with the previous traditional shallow machine learning. This concept comes from the study of artificial neural network. Multilayer perceptron is a kind of feed forward artificial neural network model, and also belong to a kind of deep learning structure, which is  through the establishment of simulation analysis of the human brain neural network to achieve the goal of learning.

The appearance of deep learning technology makes many of the pattern recognition algorithms only used in the laboratory become practical in industry.The development of deep learning accelerates the industrial transformation and is widely used in many fields such as image classification, speech recognition, target detection and computer vision.

Deep analysis of personalized needs and create customized solutions

It is known that the intelligent multimedia technology research center of the Chongqing research institute of the Chinese academy of sciences is an innovative research and development institution jointly established by the image generation and processing research office of the university of Illinois (UIUC) and the study and vision research group of the national university of Singapore (NUS).The center focuses on the theoretical and industrial application research in the field of machine vision and pattern recognition, focusing on the core key technologies such as large-scale dynamic face recognition, massive video semantic retrieval, human-machine collaboration and networked video event reasoning.

Aiming at the strategic needs of image and video information extraction and intelligent analysis in the security field, the center has independently developed a series of intelligent security products and elaborately created intelligent security system solutions for schools, customs and other enterprises and institutions.At the same time, in view of the urgent needs of the Internet industry for the in-depth analysis of image information and intelligent image processing, it provides Internet companies with a complete picture depth analysis technology solutions and creative cultural products.

In 2018, Sugon awarded Chongqing environmental-friendly intelligent technology research institute of Chinese academy of sciences, provided targeted high-performance solutions and focuses on the need of computation in the related field of artificial intelligence, deep learning for multimedia technology research center and data center,  including specially XMachine series GPU server products for the field of artificial intelligence, which can effectively satisfy the center to count the demand of the density field of artificial intelligence research.

Deep study in the field of artificial intelligence and continuously release AI computing power

As a China's leading high performance computer manufacturer, Sugon depends on its own profound experience and understanding on the application of the customer and is committed to provide users with a complete integrated machine learning/deep learning scheme of  XSharp, XMachine, XSharp + XMachine, which realize the rapid growth of user application. Sugon wins great concerns by its outstanding achievements in the field of artificial intelligence.

During the GTC China 2018 conference, Sugon officially published the brand-new XMachine series of domestic AI servers to meet the growing market demand. According to the introduction, XMachine is a series of GPU servers specially designed for deep learning, which also takes into account the high-performance computing market. "Transformers" series AI server is a GPU server product developed by Sugon itself, which also fills the gap of domestic security GPU product line and lays a solid foundation for future GPU product development.

At the same time, at the application level, the "open source platform construction and application project oriented to deep learning application" led by Sugon was successfully selected into the national development and reform commission's major project of 2018 to promote the integration development of high-performance computing and artificial intelligence and promote the innovation of deep learning application.

As an important research field in machine learning, deep learning has attracted more and more attention in recent years. At present, deep learning is gradually stepping into the application and expansion stage, and deep learning with informatization will also become a new development trend. Sugon will continue to increase the investment in hardware, applications and services based on its own technical advantages in the field of high performance computing, and continue to launch high-quality product solutions to adapt to the development trend of AI aiming to accelerate the development of the industry.

We could believe that "Jarvis" intelligent housekeeper will become a reality in the near future with the development of high performance computing, artificial intelligence and other new technologies.