Chinese voice assistant Intelligent 360 was released

Released artificial intelligence semantic open platform -- Semantic Cloud, oriented to the APP field

Smart 360 has more than 10 million users and has become a third-party voice assistant with more than 10 million users

Released ai speaker -- Smart Super Speaker (Amazon Echo released in December 2014), equipped with semantic cloud brain

Become a member of the Council of China Electronic audio Association

Released xiaozhi Butler robot, equipped with semantic cloud brain 2.0

The Smart Housekeeping robot and HTC Vive virtual reality headset won the CHINABANG Award for Innovation of the Year 2016

New upgrade semantic Cloud, released DeepBrian brain platform, new upgrade

Launch cloud brain artificial intelligence operating system blockchain-based research and project approval; Set up blockchain laboratory, based on blockchain application architecture research and development

Obtained angel investment of 32 million RMB from Jinshajiang Venture Capital, Gobi Investment and Qian Shi Investment.

1. The HTML5 version of DeepBrain's skills store is released, giving consumers the freedom to add skills to AI hardware. 2. Version 0.6 of the white paper on deep Brain chain token sales was released.

Won the first place in SMP2017 Chinese Man-machine Dialogue Evaluation of HIT University of Technology; More than 20 domestic top ai teams participated in the competition; Deep Brain Chain White paper version 1.0 released.

1. The R&D team moved to the new hongqiao Tianjie Office in Shanghai Dahongqiao Business Circle; 2. Invited to attend the Closed-door forum of 30 Chinese AI people held in Beijing and the internal seminar on Interpretation of the Notice of The State Council on the Development Plan of The New Generation of Artificial Intelligence; 3. Invited to the Digital Asset Summit.

1.10 Invited to attend "the First Global Fintech and Blockchain China Summit 2017"; 2. Completed the first-stage r&d of deep Brain chain based on NEO contract, and issued tokens; 3. Support coin charging and withdrawal. DBC assets are on-line on third-party exchanges; 4. Disclosure of quarterly key information of the project; 5. Deep brain chain promotion and manufacturer access.

Complete the development of the underlying framework and core components, complete the development of DBC AI test network, and build an internal test environment; Internal AI test users can submit AI training requirements to the DBC AI test network and complete training tasks through the DBC AI test network. DeepBrain Chain official website community task crowdsourcing system development; End the first round of global Meetup (Dublin, Hamburg, Amsterdam, San Francisco).

Complete AI user management, group management, role management, rights management development work; Complete the integration and development of DBC AI test network and blockchain system; Complete the construction and testing of DBC internal network environment; June 30th sky Net test line; Establish a Silicon Valley research and development center.

To further improve and test the blockchain network, and launch internal network testing for key users; Support the integration and deployment of multiple deep learning engines; Support model algorithm publishing and selection of published model algorithms, and release training/testing tasks after packaging data; Support DBC network monitoring and statistical analysis; AI users can perform AI calculation on the DBC network. Support DBC network security requirements; On August 8, Sky Net officially went online, DBC can be used to buy GPU computing power.

Further improve and test the blockchain network, and launch the public beta network; Further support the integration and deployment of multiple deep learning engines; Support AI user rating for task execution; Support API open to third party users; Further improve DBC network security.

Further refine and test the blockchain network, making changes based on user feedback and community needs; Support AI data trading and evaluation; AI data encryption and privacy protection.

Complete THE GPU isolation function. Design and develop new DBC message encryption function and reconstruct DBC message signature function; ECDH algorithm is used to generate one-time Share secret, and then share Secret is used for CBC-AES encrypted transmission of log messages. Integrate DBC message encryption and DBC message signing. Design, develop and complete offline authentication of training tasks, and solve the problem of free use of mining opportunities in the case of failure of Auth WEB; The DBC matrix protocol was optimized and the enable_FAST_forward option was added.

Transparent forwarding of optional parameters is supported. The current DBC intermediate node discards the fields it cannot recognize when forwarding messages. Transparent forwarding of unknown messages is supported. Reconstruct DBC source code and prepare open source documents; Supplement internal design documents of restful apis.

The standard template for creating a GPU Server is provided in the following scenarios :(a) a GPU Server uses the public IP address delivered with a Computing Node. (b) GPU Server uses Ngrok Proxy; Optimized GPU Server creation: Reduce the GPU Server creation time from several minutes to tens of seconds by: (a) Installing the standard template startup code in 2.2 on a Computing Node and copying it to each GPU Server; (b) Using standard templates will skip IPFS installation, startup, and code download; The functions of memory management, CPU management and hard disk management are developed. The GPU function is added and uninstalled to the container.

GPU container and CPU container functions are available. Automatic machine verification function is added online. After miners add machines, the system will automatically verify the availability of the machines. The whole process from deep brain chain payment network to AI computing power network to cloud platform will be automatically verified without any manual intervention. Support TensorFlow2.0 container image online, users do not need to install TensorFlow2.0, can directly use; The private web disk function has been launched, facilitating most users to upload files to the GPU server. Online image support drug molecular dynamics simulation, convenient for the majority of drug research and development workers to use GPU computing power.

Added tensorBoard support for AI users to visualize during training; Added SSH authentication mechanism of mining machine line; Cloud platform servers are hardened to prevent data loss. GPU machine classification mechanism was launched to classify machines according to memory space and hard disk space. Improve the machine rental notice information display, convenient for miners to view the machine income; A payment verification mechanism is added to prevent payment failure caused by network problems. Automatic image acquisition was completed, and a new version of image was made. The absolute value of CPU cores was added to prevent malicious mining from wasting resources. Improve the machine on-line and off-shelf mechanism; Added the hard disk restart limit, hard disk space is insufficient prompt mechanism; Improved the switching mechanism between GPU container and CPU container, which no longer depends on the size of the data in the container, greatly improving the user experience. The original 100GB of data took 1 hour to start the container, but the new version takes 3 minutes to complete; Script startup mechanism is added in distributed computing power network layer. A new version of the image is created to unify the Python execution environment and the jupyter and private web disk file directories.

Added automatic recovery mechanism in case of failed switchover between GPU container and CPU container. Added the disconnection automatic recovery mechanism of Jupyter and private web disk Ngrok; Added the automatic recovery mechanism for the abnormal termination of SSH, JUPyter, and private web disk processes. Developed and produced brand new images of TerSORFlow2.1, TersorFlow2.0, TensorFlow1.14, PyTORch1.2, PyTorch1.4, PyTorch1.5, MXNet1.5, MXNet1.6; A new update to deep learning mirrors has removed a lot of junk content and made the mirrors more lightweight. Added mxnet1.5, mxnet1.6, tensorflow2.1, pytorch1.5, the latest molecular dynamics environment support; Each image adds a corresponding version of ipykernel kernel to facilitate the use of JUPyter. The virtual environment of the image version is automatically activated to facilitate the use of the image. Unified virtual environment jupyter and terminal access under the Python version; Upgrade Ubuntu to 18.04 Image preinstalled opencV compiled version, greatly improve the use of OpencV experience; Added docker automatic detection and restart function; Detailed server usage documentation and help documentation have been added; The response mechanism of AI computing network is optimized. The deep learning image and DBC network are upgraded, and the image is preinstalled with opencV compiled version, which greatly improves the experience of using OpencV. Added PIP default download source for Tsinghua source; The response mechanism of AI computing network is optimized. Cuda9 support is added by default; Filecoin Snark computing image support API access for FILecoin seal C1 and C2 computing support; Added support for Filecoin Snark GPU computing.

The discount function of monthly, quarterly and annual package has been added. Added support for the whole rent function; [Fixed] Added AI reasoning area, additional open ports are available for machines that rent this area. The website interface layout has been redesigned to make the classification clearer; Increased the DBC super node rental function, want to participate in THE DBC super node, do not need to buy their own machine, only need to pledge a certain number of DBC can rent THE DBC network machine, participate in the node block; Added the pledge function. If you want to enter the high stability AI training area, you need the machine to run steadily for at least 144 hours and pledge DBC. Added the function of mounting machine and removing machine; The GPU acceleration algorithm is optimized. Optimized the cloud platform wallet and added the gas function; Research the combination of privacy computing and DBC network; Optimized the node dynamic expansion function of DBC network; Deep learning images for DeepShare courses and papers have been updated; In order to cope with the continued increase in the DBC network machine, further optimize the DBC node dynamic expansion function; Put AIM machines online. Now all machines are online and can be delivered to AIM buyers at any time; Further optimize the MESSAGE forwarding function of DBC network node; To meet the needs of cloud gaming and rendering users, develop a scheduling engine based on GPU virtual machine; Modifying substrate based web wallet; Nearly 30 GPU cloud platforms based on deep brain chain are supported online.

DBC main network public test phase one start; Develop online award contracts for compute nodes; Develop compute node rental and destruction contracts.

DBC main network public test phase two start; Computing node online award contract Test Phase I; DBC main network officially online, start the token exchange DBC; The online award contract for compute nodes is launched.

The compute node lease and destruction contract officially went online, and THE DBC began to enter the era of deflation.

Upgrade and improve smart contracts; Support AI training and reasoning, cloud gaming, graphics rendering, ETH2.0 and other multi-scene customer advanced customization requirements.