Inspur Blog

Meeting customization demands of the growing cloud data center with Inspur JDM

The explosive growth of cloud computing applications like e-commerce, big data, mobile internet, IoT and other transformative information technologies call for large-scale customization in the cloud data center. Traditional whitebox OEM and client-branded ODM production models are no longer able to support the increasingly fast and complex customization needs of enterprises.

Over the course of 20 years in the computing hardware industry, Inspur refined the JDM (Joint Design Manufacturing) business model. Based on the integration with the user’s industrial process, this model is aimed at the specific business of the user, handling demand, R&D, production and delivery, supply chain, designing and customizing according to client needs, and solving a very simple but critical challenge – develop products that users need and deploy them to the user data center in the shortest time and within cost constraints.

Inspur’s JDM model consists of 3 key processes: close customer collaboration; design, R&D and development; intelligent manufacturing to enable agile production and fast delivery.

1) Extensive Customer Collaboration

 IDC analysts believe that the customization model requires more extensive collaboration between server vendors and end users, putting higher demands on the vendor’s R&D and resource integration capabilities. To meet these demands, Inspur’s new product development line is updated once a year, cutting new product development time from 1.5 years to 9 months, and R&D to supply can take as little as 3 months.

Development Team

Each Inspur client is assigned a user-specific “iron triangle” team of product manager (demand docking), architect (architecture design) and project manager (product implementation). Product definition, product architecture and R&D plans can be completed in the shortest time to fulfill user needs.

The Inspur user team visits clients regularly to understand business and user requirements. For some small-to-medium CSPs, Inspur actively participates in the user infrastructure planning and engages user engineers throughout the process of product development. Inspur can adjust the development process of each product in time according to user needs, eliminating unnecessary processes.

2) Design and R&D

Global R&D Capabilities

With the support of eight global R&D centers including Silicon Valley in the United States, Inspur has the ability to provide customized solutions for all CSP users around the world, and has independently developed key application hosts such as Tissot K1 and M13. Featuring leading system design and optimization capabilities, Inspur is one of few industry vendors with architecture and high-speed interconnect chip R&D capabilities.

3) Intelligent Manufacturing

To answer the need for cloud server manufacturing to shift from low-efficiency traditional production lines, Inspur built the industry’s first intelligent flexible production base. Two intelligent flexible production lines and 10 parallel production units can simultaneously produce 10 different types of cloud server nodes which are assembled in a unified heavy assembly line.

Data-Driven Automated Production

Six core intelligent information systems including ERP, MES, and WMS, control the automated operation of assembly and transport robots, and intelligent stereoscopic warehouses capable of smart scheduling, storage, transportation and tracking.

The ERP system enables the collection and analysis of data from each component of production, using both chip-based automatic data acquisition and manual scanning, to provide product traceability, supplier evaluation, and strict quality control.

With data-driven automation and agile, efficient parallel production, Inspur is able to meet the dynamic demands for customized cloud computing and big data products, like deploying 10,000 nodes a day in Baidu’s Yangquan data center.

Testing and Quality Assurance

Inspur’s product quality control spans R&D, strict component procurement, production process control, and strict downline quality inspection. It is also possible to pass the road spectrum test and strictly control the transportation and implementation of the products.

Eight R&D labs test product feasibility: climate environment, mechanical environment, failure analysis, structural heat dissipation, electromagnetic compatibility, precision measurement, system verification, and data center. Other component tests are conducted to resolve problems in the entire production chain.

Quality control extends to transportation and delivery, where each device is shipped with special electronic tags that transmits the geographical location and conditions in real time to Inspur ERP system, viewable through Inspur engineers through a mobile app.

The future of JDM

The JDM model enables Inspur to massively deploy to major data centers across the globe. Inspur’s network spans 113 countries and regions, 8 global R&D centers, 6 global production centers, propelling Inspur’s server shipments to third largest in the world in 2017 according to IDC data. To date, Inspur has made large scale server deployments to six out of the world’s top ten CSPs.

Cloud servers will lay the foundation for major innovations in emerging areas like AI, reducing cost pressures and presenting limitless opportunities for enterprises. As cloud computing grows in importance in the development of future technologies, the JDM operational model will continue to play a critical support those demands.