eyeM5

eyeM5 5500/550 FPGA Mining Machine

  • Value Proposition
  • Comparison
  • Photos
Value Proposition
  • Energy saving. The power consumption of FPGA mining machine is20 times more efficient than that of GPU (FPGA: 80W, GPU:1,650W)
  • Unmanned maintenance management in the AI Cloud Data Center
  • More than one-year warranty for eyeball55 chip, differentiated from the half year warranty of GPU mining machine
Comparison

between eyeball55 FPGA and GPU ETH Mining Machine

nVidia/AMD GPU Mining Machine

  • Function: ETH mining
  • Power Consumption: 1650W*100 GPUs*24hours/day=3,960,000W/day (3,960° electricity/day)
  • Computing Power:285-320Mhash/s
  • If the annual yield is 8 ETH Coins(by nVidia P104 N)
  • Gross Profit: 8/year * current market value of ETH – annual electricity cost (3,960°/day *365 *$/°)

eyeM5 5500FPGA Mining Machine

  • Data showed on the mining log is based on the ETH algorithm + AI forecasting model.
  • Power Consumption: 80W*100 FPGA PCB*24 hours/day = 192,000W/day (192° electricity/day)
  • Computing Power: 270-300Mhash/s
  • If the annual yield is 8 ETH Coins
  • Gross Profit: 8/year * current market value of ETH – annual electricity cost (192°/day *365 *$/°)
  • Gross Profit Comparison by power consumption: More than 20 times of GPU 3,960°electricity/day to FPGA 192°electricity/day

Rate of Return of FPGA vs GPU

  • The rate of return of eyeM5 FPGA 5500/550 is 20 times more than that of nVidia P104 GPU mining machine.
Photos

There are two models of eyeM5 FPGA mining machine, one is eyeM5 FPGA 5500 with 4 FPGA PCBs mounted in one unit, the other eyeM5 FPGA 550 with 1 FPGA PCB mounted in one unit.


Source: Eyeball FinTech Co., Ltd.


eyeM5 550 AIoT Computer / Platform

eyeM5 550 AIOT computer can be used as AI personal computer as well as developer’s platform. As shown below the functions of IoT applications, deep learning, algorithm, parallel operation are all integrated in the platform for developers with friendly user interface. eyeM5 AIoT Development Platform is mounted with AI API, Hyperledger API, R/Phyton/Go languages to enable developers to collect various types of data for edge computing. It also adopts IPFS protocol to encrypt and transmit data to the cloud environment. eyeM5 AIoT platform is developed for domains covering smart finance, precisive agriculture, smart city, smart manufacturing, smart logistics, etc., and allows developers to deploy algorithm easily for specialized application scenarios.


Source: Eyeball FinTech Co., Ltd.


eyeM5 550 AIoT personal computer is a desktop device as shown below. The monitor may be selected by the client.

eyeM5 550 AIoT Personal Computer / Source: Eyeball FinTech Co., Ltd.


Source: Supplier


eyeM5 5500~9500 edgeX HPC

Servers Used in AI inside60 Data Service Platform

  • Technology
  • Solution
  • Fuctions

eyeM5 5500~9500 edgeX HPC Server
Source: Eyeball FinTech Co., Ltd.
eyeball55 AI chip is a core of

eyeM5 Hyper-leger technology used to encrypt distributed ledgers and to develop into edgeX HPC servers with codes of eyeM5 5500~9500. eyeM5 edgeX HPC series of servers empower clients’ existing database with AI modules to transform the database into intelligent ones. The eyeM5 edgeX HPC solution can also ensure security when making point-to-point transaction and data transmission on the Blockchain platform. Its architectural features of irreversibility and unchangeability of data can help fulfill the goal of data security and certificate protection.


Source: Eyeball FinTech Co., Ltd.
Solution

eyeM5 5500~5900 edgeX HPC series of servers combine and integrate technologies of supply chain partners with FPGA AI chip, PCB, and server to form an ecosystem supporting innovative applications in various business domains. The diagram below illustrates a scenario of digital financial application using eyeM5 5500~9500 servers branded as “AI inside60”.

Fuctions

AI inside60 Engine is a cluster of eyeM5 FPGA based AI servers consisting of the following series of servers, each of which has its own functions and service provisioning as shown below.

  • eyeM5 5500 Big Data Aggregator
    A server adopts IPFS (InterPlanetary File System) protocol for managing distributed databases to store and share data amongst different file systems. With the IPFS protocol eyeM5 5500 server incorporates two features: one is content-addressing, the other not tampering. Due to multiple hash verification the former enables the contents on the web to be uniquely identified using a global namespace across various computing devices. The latter is because all the contents are also hash verified, any contents which seem to be changed or revised will be recognized and picked out to ensure data/contents integrity.
  • eyeM5 6500 Memory Database Cluster
    A server specializes in managing “Time Series Data” which needs synchronous handling of IO R/W and data computing and storage. The features of high speed and low latency are provided by eyeM5 6500 server.
  • eyeM5 7500 Data Access Rights Management Server
    A server provides a function of data access rights management for data security assurance. During the modeling process for sensitive big data the access rights are specified based on the nature of data/contents and personnel’s position/responsibility in order to fulfill the requirement of data security and data management. Besides, a secure billing module on data stream is incorporated and the system security and stability strengthened. The data especially for AI training will be encrypted and transmitted.
  • eyeM5 8500 AI TensorFlow Server
    eyeball55 AI chip is burned with a Digital Signal Processer (DSP) on Recurrent Neural Network (RNN) in the FPGA Logic Cell, the eyeM5 8500 AI TensorFlow server manages and provides up to 36 IP codes to identify including general image, facial image, passerby’s posture, dynamic video, Advanced Driver Assistance System/Autonomous Driving, ADAS/AD, etc.
  • eyeM5 9500 Hyperledger Server
    Burned with eyeball55 AI chip the server is a middleware for secure blockchain applications, which enables the smart contract to access the external resources of the critical chain such as database.


Ready to get Started?

Contact Us Now!