The application of high-performance and distributed computing in scientific practice has become more importance, among the most available platforms such as clusters, grids and cloud systems. These infrastructures are now undergoing many changed due to the integration of core technologies, providing speed improvements for selected compute kernels. As the distributed and high-performance computing is becoming more heterogeneous and hierarchical, complexity in programming is increased. Further, these complexities arise due to urgent desire for scalability and issues like data distribution, heterogeneity in software and hardware availability. These issues force scientists into simultaneous use of multiple platforms (e.g. clusters, grids and clouds used concurrently).

6.1 Jungle Computing

Jungle Computing is a distributed computing paradigm. It simply emerged out of the plethora of distributed resources available. A Jungle Computing System consists of all compute resources available to end-users, which includes clusters, clouds, grids, desktop grids, supercomputers, as well as stand-alone machines and even mobile devices.

There are several reasons for using Jungle Computing Systems. Firstly, an application may require more compute power than available in any one system a user has access to. Secondly, different parts of an application may have different computational requirements, with no single system that meets all requirements.

From a high-level view, all resources in a Jungle Computing System are in some way equal, all consisting of some amount of processing power, memory and possibly storage. End-users perceive these resources as just that: a compute resource to run their application on. Whether this resource is located in a remote cloud or located down the hall in a cluster, is of no interest to an end-user, as long as his or her application runs effectively. Despite this similarity of resources, a Jungle Computing System is highly heterogeneous. Resources differ in basic properties such as processor architecture, amount of memory and performance. As there is no central administration of these unrelated systems, installed software such as compilers and libraries will also differ.

For example, where a stand-alone machine is usually permanently available, a grid resource will have to be reserved, while a cloud requires a credit card to gain access. Also, the middleware used to access a resource differs greatly because of using different interfaces.

The heterogeneity of Jungle Computing Systems makes it hard to run applications on multiple resources. For each used resource, the application may have to be re-compiled or even partially re-written, to handle the changes in software and hardware available. Moreover, for each resource, a different middleware interface may be available, requiring different middleware client software. Once an application has been successfully started in a Jungle, another aspect that hinders usage of Jungle Computing Systems is the lack of connectivity between resources.

6.2 Jungle Computing Systems

When grid computing was introduced over a decade ago, its foremost visionary aim was to provide efficient and transparent (i.e. easy-to-use) wall-socket computing over a distributed set of resources. Since then, many other distributed computing paradigms have been introduced, including peer-to-peer computing, volunteer computing and more recently cloud computing. These paradigms all share many of the goals of grid computing, eventually aiming to provide end-users with access to distributed resources (ultimately even at a world-wide scale) with as little effort as possible.

These new distributed computing paradigms have led to a diverse collection of resources available to research scientists, which include stand-alone machines, cluster systems, grids, clouds, desktop grids, etc.

With clusters, grids and clouds thus being equipped with multi-core processors and many-core ‘add-ons’, systems available to scientists are becoming increasingly hard to program and use. Despite the fact that the programming and efficient use of many-cores is known to be hard, this is not the only problem. With the increasing heterogeneity of the underlying hardware, the efficient mapping of computational problems onto the ‘bare metal’ has become vastly more complex. Now more than ever, programmers must be aware of the potential for parallelism at all levels of granularity.

But the problem is even more severe. Given the ever increasing desire for speed and scalability in many scientific research domains, the use of a single high-performance computing platform is often not sufficient. The need to access multiple platforms concurrently from within a single application often is due to the impossibility of reserving a sufficient number of compute nodes at once in a single multi-user system. Moreover, additional issues such as the distributed nature of end-users, simultaneously comprising any number of clusters, grids, clouds and other compute platforms.

Summary

For every new technology several research frontiers are to be exploited. So in Cloud computing. Topics of such a kind are effective data protection in Internet clouds, innovative applications on the clouds, data centers and the Internet of things (IoT). The material in the Chapter 1 deals with future trends of cloud computing, next-generation services related to cloud computing are explained. With the emergence of the ‘mobile cloud’, more and more productivity applications residing on mobile devices are developed. Chapter 2 details how mobile and cloud computing can be combined and also exploits its key requirements. It is time to design and build computing systems capable of running adjusting to different circumstances and using their resources to handle most efficiently the workloads we put upon them. These autonomic systems combined with cloud computing is called as Autonomic Cloud computing is discussed in Chapter 3. With Web 2.0&3.0, Internet multimedia is emerging as services. To provide rich services in media, multimedia computing became a promising technology to generate, edit, process, search media contents which includes audio, image, video and graphics. Chapter 4 presents principal concepts of multimedia cloud computing. Energy efficiency is an important aspect IT field. Energy consumption, resource utilization and performances of workloads in Cloud are dealt in Chapter 5. We need a platform which will need speed and scalability in everyday scientific practice and the resources employed by end-users are often more diverse than those contained in a single cluster, grid, or cloud system. Jungle Computing is explained in Chapter 6.