WISE-STACK Fully Integrated Edge Intelligence Private Cloud

Enjoy the Experience of a Public Cloud, with the Confidentiality of a Private Cloud

WISE-PaaS 3.0 On-Premise Deployment

WISE-STACK is Advantech’s private cloud offering with an on-premise deployment method. The fully integrated system offers the complete cloud experience and development toolset. End-to-end security features and a user defined access pathway ensures security and privacy of data and AI models. WISE-STACK is designed and tested to host AIoT applications with high availability, elastic expansion, and high data throughput characteristics. All WISE-PaaS services offer seamless integration with Advantech edge devices. Companies can easily complete their IoT initiatives by combining Advantech’s edge computing products with WISE-STACK.

System Architecture

private cloud architecture

Feature Highlights

The Benefits of WISE-STACK: Fast System Integration and High Data Throughput

wspc feature-1

Validated performance, elastic expansion, and high availability for industrial AIoT

wspc feature-2

Seamless integration with Advantech edge devices

wspc feature-3

Complete data and AI model privacy

WISE-STACK Configuration

Advantech provides standardized, on-premise cloud configuration to satisfy large computation and high availability needs. Starting configuration consists of six server nodes and one optional analytic node with GPU.

wise stack configuration
wise stack configuration-01

Controller Node: Controls operation and coordination of servers to achieve load balancing and high data throughput. Three controller nodes guarantee high availability for operational continuity, even when a server goes down. Users can perform hot-swap to maintain operation and ensure the performance of their private cloud.

HCI Node: Pure, hyper-converged node provides high computation power and storage space. The configuration allows businesses to setup their private clouds with a reduced number of servers in the initial stage.

Analytic Node: WISE-PaaS architecture incorporates Kubernetes – utilizing the power of NVDIA Tesla P4 graphic cards – to shorten the training period for deep model learning.