Expectations of computing technology are growing. Whether those expectations come from commercial business ventures or public service, such as government, the requirement that computing technology delivers increasing value at lower costs is a consistent theme.
IBM has adopted the concept of a smarter planet to connect the role of technology with the needs of business and government to solve problems in markedly different ways. The planet is more connected then ever. With the Internet, IT is now able to connect to the real world, and more able to access real-world data for deeper analytics. With connections to customers and suppliers, and devices such as meters, microphones, etc., IT can create more sophisticated applications (i.e., smart applications) that can address problems in new ways.
Smart applications will enjoy connections to improve processes that integrate more deeply and automate more robustly. Process improvement will come from the ability to connect to more business process constituents.
There are limitless possibilities for smart applications. For example, think of:• Metro microphone towers wired to 911 call centers providing event recognition (such as gun shots) and triangulation capabilities integrated with incident reporting systems. Consider how parolee systems enable law enforcement officers to arrive armed with a belt full of equipment and a laptop full of highly pertinent information.
- Metro microphone towers wired to 911 call centers providing event recognition (such as gun shots) and triangulation capabilities integrated with incident reporting systems. Consider how parolee systems enable law enforcement officers to arrive armed with a belt full of equipment and a laptop full of highly pertinent information.
- A smart little meter on the side of your house reporting to your utility provider your use of its services every 15 minutes rather than once a month. Think of the richness of the data now afforded your provider so they might better serve you and your community.
- Goods flowing from suppliers in raw material form, progressing through manufacturing into finished goods form, and then through the supply chain to wholesalers and retailers, and finally to consumers where all goods are tagged and tracked with Radio Frequency Identification (RFID)-based instrumentation, promoting deep integration of tightly coordinated systems.
Extracting Business Value From Information Stores
Instrumented new processes will collect data for insertion into operational processes and for gathering into informational data stores. When combined with existing data warehouses or data marts, these informational data stores will offer the raw material for a well-rounded, analytically robust view of the process space. The rich data content, combined with sophisticated reporting and analytic software, will support increasingly complex analysis, yielding more effective and timely business decisions.
Affording Systems for a Smarter Planet
But sophisticated software comes at a cost in terms of execution. The servers running smarter planet applications must execute more functions in the scope of each transaction. Transactions are what it’s all about! Servers must deliver acceptable, predictable transaction response times (seconds per transaction) while scaling to deliver the throughput required for business success (transactions per second) in an affordable fashion. In short, servers and systems must remain affordable even in the face of rapidly growing software sophistication.
Throwing more computing resources at the problem will only go so far. More sophisticated computational capabilities are needed. Application domain computing and high-performance computing are two examples of computational models that, where applicable, drive large price-performance gains over general-purpose computing (i.e., just running software). The perfect storm comes from harnessing all three computing models under the same covers. What IT requires is a powerful, general-purpose computing system that promotes the integration of application domain computing elements (i.e., appliances) as well as high-performance computing elements (i.e., parallel processing) in service to the workloads being executed (i.e., in accord with the needs of the software to perform and to scale).
Composing a Smarter Server