Harnessing and distributing more power throughout the enterprise
Grid computing harnesses the power of many CPUs across a network to help process problems too large for standalone machines. Autonomic refers to computer systems and networks that configure themselves to changing conditions and are self-healing when things go awry. Such concepts are rapidly emerging as significant forces in commercial computing, similar to the way the Internet and the World Wide Web emerged from research and academia.
The IBM Research Services Grid and Autonomic Computing micropractice applies cutting-edge models, software, designs and expertise to help quickly and efficiently evaluate, design, pilot and optimize grid and autonomic capabilities in distributed computing systems.
This cutting-edge micropractice employs IBM Research expertise in high performance computing (HPC), parallel computing, data management, transaction processing and modeling to help companies infuse their distributed systems with grid and autonomic technologies, to help them understand and optimize their computing resources and applications, and to help make their systems more resilient.
- For a financial company, this may mean providing grid infrastructure and solutions to speed up applications such as portfolio optimization, risk analysis and Monte Carlo simulations. It may also provide autonomic capability for sharing computing resources between applications, such as portfolio optimization and transactional applications, with dynamic resource allocation to optimize resources.
- For an industrial company, this may mean grid-enabling distributed computing resources to make analysis applications, such as engineering design, seismic and other complex analyses, faster. It may also provide autonomic capability for on demand computations, either within an enterprise or through IBM’s on demand supercomputing offerings.
- For a life sciences company, it may mean accelerating applications used in bioinformatics and molecular modeling.
Specialties
Specialties within this micropractice focus on helping businesses evaluate, design and optimize their distributed systems and infuse grid and autonomic capabilities to better meet their business goals:
- Autonomic architecture helps businesses create a computing environment that is structured to promote autonomic computing. This process typically begins with an autonomic computing assessment, offered by IBM Global Technology Services, to gain a more detailed understanding of the computing environment. For example, this analysis determines where monitoring takes place and where there are opportunities to consolidate monitoring and reporting.
- Autonomic infrastructure focuses on building an autonomic infrastructure that supports automating costly IT tasks. Examples of these tasks are log analysis, problem determination and initial configuration for servers, networks and storage. IBM helps companies deploy appropriate instrumentations, assist with the development of workflows to effect actions and asses operational autonomic infrastructures.
- Autonomic tools assist a company with developing skills and applying various tools used to construct and deploy the autonomic infrastructure. The tools relate to various aspects of autonomic computing, including policy, sensor development (e.g., log access), and construction of information models. One near-term benefit for businesses is the use of the common base event (CBE) so as to gain access to new technologies for event reporting and automation. IBM helps companies inventory their existing infrastructure, build adaptors that facilitate the exploitation of autonomic capabilities and co-develop measurement probes that are representative of end-user interactions so that quality of service can be measured.
- Data grid specialists help companies design and build distributed data management across grids. This includes data virtualization or transparent access to data sources. For example, the applications need not be aware of the location or schema of the data, and replicas of data can be used to provide quality of services goals. Areas of expertise include data distribution, federated queries, data integration and distributed file systems.
- Grid infrastructure and tools specialists focus on providing tools and components that can be used across different types of grid environments for the evaluation, design, implementation and operation of processing, data and transactional grids. These tools and infrastructures can be applied to all of the evaluation, design and operational phases of grid enablement.
- Processing grid researchers focus on applying IBM's long history and technical leadership in HPC computing to help solve numerically intensive computing problems. Examples of such problems are common throughout commercial research and development labs in the areas of financial services (Monte Carlo simulations), automotive design (finite element analysis), aerospace (computational fluid dynamics), pharmaceutical research (computational chemistry), as well as in national government labs and universities.
- The transactional grid specialty focuses on sharing resources among transactional applications in clustered systems and in multi-site grids. This specialty can help companies estimate the return on investment of a transactional grid for their applications, and provide expertise in the design, implementation and customization of transactional grids for applications and scenarios. It also can be used in combination with the autonomic infrastructure and processing grid specialties in cases where a company can share resources between transactional and computer intensive applications.
The specialties listed do not represent a comprehensive list of IBM Research capabilities. Business leaders with ideas or custom solutions in mind, please contact IBM Research Services today.