Prepared for Enterprise 2 0 Mashup Summit September

Prepared for Enterprise 2. 0 Mashup Summit September 28, 2007 Enterprise Mashups in Outsourced Manufacturing Mashing your Shipments and Processes Serus Corporation Jeffrey Risberg VP of Research and Development

Introduction • “Outsourced Manufacturing” defined: – The use of contract manufacturers, semiconductor fabs, assembly houses, and logistic organizations outside your business • Significantly increased in last 10 years – By 2010, analysts estimate that $1 trillion of manufacturing will be outsourced – Cisco, for instance, never touches its inventory – There are 238 fabless semiconductor firms in Silicon Valley alone • As described in “The World is Flat” by Friedman, outsourced manufacturing has generated a new set of challenges and complexities • Over 50% of the information needed for today’s manufacturing resides outside an individual company – It is with the company’s Partners, Customers, Vendors, and in the Public Ecosystem

Ideal Case for Enterprise Mashups The challenge is to deliver the Right Information, to the Right Place at the Right Time. Required is a Knowledge Multiplier for Outsourced Manufacturing Operations

Operations Management - Challenges “Global Outsourcing” has generated a unique set of operational challenges: – – Lack of Visibility to Financial Exposure Lack of Consistent Operational Data Lack of Execution Management with Partners Difficulty in Collaborative Decision Making with Partners

Content Used in Outsourced Manufacturing Private Product Manufacturing Specifications Note: Customer Information are not part of Forecast and Order Information Starbucks locations the data set! Activity records Public Transportation links Weather/Geography Currency Information Compliance Information (Ro. HS) Ecosystem Scores, Behavior, Prior Experience

Architecture of the Serus Solution Mashup Server

Architecture of the Serus Solution Traditional Scope of Mashup Servers

Architecture of the Serus Solution Processes Serus Definition of Mashup Servers Data

Example of Mashups Applied to Manufacturing

Knowledge with Ability to Act

Partial List of Vendors/Partners already supported by Mashup Server Fab/Assembly Contract Mfg Distributors

Enterprise 2. 0 Definitions and Implementation Andrew Mc. Afee: SLATES • • Serus: CONTROLS Collaboration: sharing of quantitative information Options generation: system generated suggestions Notifications: alerting Tags and content: marking most relevant content, allowing users to mark Real-time data: dynamic fetching (JDBC, FTP, HTTP) Open system: use of standards for data fetching (WS, XML) Links: navigate through a cohesive data model Scenarios: user-controlled “what-ifs” Serus Implementation • Java-based web application, supported by multiple back-end processes

Foundations of Architecture

Key Architecture Components From inception, the Serus Architecture has been built on the following four cornerstones: • SOA – Service Oriented Architecture – Serus application invokes and orchestrates services using remote protocols such as Web Services, RMI, JMS • EII – Enterprise Information Integration – Serus mashup server extracts, transforms and loads from multiple data sources – Serus mashup server resolves information conflicts between sources • BPM – Business Process Management – Processes can be defined using workflow, and rule sets – Processes can carry out validation and be audited • Scenario-based Analytics – Analytics such as KPIs – Analytics evaluated within what-if scenarios

Solution Architecture

Currency Information Enterprise Mashup Example: Order Resolution Compliance Information Common locations table Cisco Dock 1 Solectron Facility 1 Avnet Recv 1 Spansion Dock 1 TSMC Ship 2 Customer 1 Multi-tenant Customer 2 data model Weather Information Geographic Information Weather Forecast: London – ptly cloudy, high 80, low 6

AMD Case Study • $7. 4 Billion Dollar Semiconductor Supplier • Global Manufacturing; Captive and Outsourced • Fragmented Manufacturing Specification • Complex Bill of Materials and Supply Chain • Manual Reconciliation between Mfg and Business Systems

AMD/INCA System Data Flow High Frequency Fault Tolerant B 2 B data transmission platform Standardized WIP formats across all sources resulting in improved quality of data and reduces the number of transaction errors Standardized Product and Manufacturing Specification format to target systems for efficient operations management Direct B 2 B transaction based connections to many major suppliers.

AMD Case Study Benefits Acknowledged: Reduced by 50% the time to close the quarter end financial records Reduced variances in inventory valuations to less than 1% Provided the “System of Record” SOX compliant Inventory Records Integrated Manufacturing specifications with execution system eliminated costly manual errors and annual scrap costs Eliminated $4 M annual support cost managing disparate legacy systems Provided support for new manufacturing complexities and business rules explosions Taiwan Dresden, Germany Sunnyvale, CA Austin, TX Penang Singapore

Conclusion • Enterprise Mashups are Critical to Outsourced Operations Management Look us up at www. serus. com