Based on the powerful big data storage and computing capabilities of cloud computing platforms, it is an important idea for large companies to cut into the field of artificial intelligence. Ma Huateng once said: "Artificial intelligence, the Internet of things, and even the future of unmanned, robots, etc., its back-end core must have a brain in the cloud." For Tencent, Ma Huateng mentioned the cloud brain is It is built on top of Tencent Cloud Services.
Following the launch of the FPGA cloud server in January of this year, the deep learning platform DI-X platform has become Tencent’s another major move in the field of artificial intelligence. It is reported that, combined with the existing business integration and division, Tencent has built a full-stack product service system in the AI ​​field, including IaaS infrastructure, AI platform services, AI basic services, AI application services, and vertical industry solutions.
Tencent Shenzhen's fast-paced product pushes the new rhythm, echoes the anxiety of the artificial intelligence during the Ma Jingteng Beijing two sessions, and the recent Tencent AI Lab... Let us gradually feel: Tencent, once restrained and low-key, seems to be ready Step into the deep water area of ​​artificial intelligence.
The initial olfactory AI pattern is undermined. Zhiwete interviewed Tencent T4 experts, Tencent cloud machine learning platform research and development head Andy, Tencent cloud big data product director Ben, Tencent cloud expert engineer Burt, etc., explored Tencent cloud for Tencent artificial intelligence underlying hair More details and development ideas for force.
First, the role of Tencent Cloud is being amplified
Wisdom: Tencent Cloud's current user size and market share?
More than one million users. There is no authoritative research data on market share.
Smart things: how does Tencent Cloud cooperate with AI Lab, excellent picture and other R&D departments?
The main method is productization. Tencent Cloud will closely communicate with Tencent's R&D team and form a joint team to continuously export Tencent’s world-leading technical capabilities to users.
Smart things: Tencent Cloud's products and services in the past two years are biased toward artificial intelligence applications. What are the intentions and logic of such trends?
Artificial intelligence, Internet of Things, and even the future of drones, robots, etc., are backed by the brain of a certain cloud. We firmly believe that the future of the main traditional industries, the use of Internet technology, the use of artificial intelligence in the cloud to handle big data. Although it is still in primary application at this stage and cannot achieve a full-scale artificial intelligence approach, the gradual transition to the future must be in the direction of artificial intelligence and big data.
Smart things: With the increase of products and services, how does Tencent Cloud enhance the security of user data?
In view of the balance between big data collection and personal privacy, we believe that on the one hand, when collecting big data, we should not collect as much personal privacy information as possible. We should collect only some necessary or even desensitized data, and we must do one-way verification. At the same time, it is recommended that enterprises adopt decentralized storage of big data and adopt a one-way encrypted storage mechanism that is user-oriented, so as to prevent internal personnel from abusing their power to steal full amount of data.
Tencent Cloud is a window for Tencent technology opening to the outside world. In terms of basic security, Tencent Cloud provides a complete set of mature network security architectures, including multiple protection mechanisms such as firewalls, distributed protection, intrusion prevention, and web application protection, to cope with various Internet Threatened. In 2016, Tencent Cloud was awarded the CSA-STAR Cloud Security Certification by DNV GL, an international authoritative certification and assessment agency. At the same time, Tencent Cloud obtained the UKAS International Recognition for Information Security and CNAS Domestic Approval. For different scenarios, Tencent Cloud has also intensively released special security solutions for applications such as live video, e-commerce, finance, and games.
Wisdom: Tencent Cloud What external industry applications can be shared?
Take the cooperation between us and Didi as an example. Didi trips are an attempt by Tencent Cloud to export technology to external companies. Dripping trips in the early stages of development have seen a surge of users in a short period of time. At that time, it was believed that the speed of expansion and the speed at which the machines were purchased by Didi Chuxing had completely failed to meet the development of the business. We will relocate the entire system to Tencent Cloud, and the back-end system of Didi's trip will be able to safely carry dozens of times of business growth within a few days.
Wisdom: Where did the AI ​​application service opened at the end of last year fall?
Taking Cloud Image as an example, it provides developers with services such as intelligent image recognition, image content recognition, face recognition, OCR recognition, etc. It can also provide customized image recognition services according to the needs of enterprises; Provide flexible image editing services, such as cropping, compression, watermarking, etc., to meet your needs for various business scenarios.
Second, the development path of deep learning platform
Wisdom stuff: What is the original intention of the DI-X project?
Like many of Tencent Cloud's superior products and capabilities, the DI-X platform has undergone an output process from the inside out. With the development of machine learning and artificial intelligence, Tencent needs a platform internally to support the urgent needs of internal algorithm engineers and data scientists. DI-X has emerged as the times require, and it has become a major machine learning platform with tens of thousands of machine learning task flow every day. On top of that, various algorithms continue to run and various models are trained.
Wisdom: What tests have DI-X done internally?
After more than a year of actual combat, the application of products such as wide-point communication, mobile phone housekeeping, application treasure, and daily express, has continued to mature. At the same time, with the increase in external demand, Tencent Cloud decided to open up this capability.
Wisdom: Which team leads the DI-X project? Team size?
The project is made up of SNG Tencent Cloud's data product team and the TEG Data Platform's machine learning team. There are only a few people. We believe that small teams are more suitable for rapid iterations, and the products are still young and need rapid growth.
Wisdom: What machine learning algorithms are compatible with the DI-X platform?
Users both support the use of integrated machine learning algorithms and can submit their own algorithms on various machine learning components in the industry
Wisdom: How does the algorithm and model on the platform update iteratively, and how does it interface with the latest technology algorithms?
In subsequent versions, we will optimize in two ways: One will add other machine learning components, including Spark, xgBoost, and even Pyhon, R. We will add them one by one to make DI-X more complete. Secondly, the characteristics of deep learning will be used for further product optimization, making it more competitive in terms of hyperparametric adjustment of the model and visualization of results.
However, the specific plan depends on the arrangement of the product manager and feedback from the user.
1 2 Next>Lifepo4 Lithium Battery,26650 Battery,3.2V Lifepo4 Lithium-Ion Battery ,26650-3300Mah Battery
Henan Xintaihang Power Source Co.,Ltd , https://www.taihangbattery.com