Electricity: electrical systems and devices – Housing or mounting assemblies with diverse electrical... – For electronic systems and devices
Reexamination Certificate
2000-09-28
2002-10-22
Talbott, David L. (Department: 2827)
Electricity: electrical systems and devices
Housing or mounting assemblies with diverse electrical...
For electronic systems and devices
C361S731000, C361S733000, C361S735000, C439S928100
Reexamination Certificate
active
06469901
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates, generally, to scalable electronic systems, and, in preferred embodiments, to systems and methods for cartridge-based, geometry-variant, scalable electronic systems.
2. Description of the Related Art
Over the last few years there has been tremendous growth in the awareness of, and the desire to utilize, on-line resources including the Internet and the World Wide Web. New on-line users are jumping in with enthusiasm and high expectations based upon the promises of cyberspace. Business has rushed in as well, with major media companies and publishers, as well as novice entrepreneurs, setting up and championing their own web sites.
The Internet, as a digital resource, is now established in many parts of the world, and is increasingly viewed as an essential utility such as water or electrical power. Furthermore, the global demand for high speed transmission and manipulation of increasingly complex data is unlikely to wane in the foreseeable future. Individuals, corporations, universities, and government agencies like the Pentagon are demanding increased communications speed and computing power to cope with the greater volume of data and the increased complexity of data handling requirements, and will likely purchase as much communication speed and computing power as they can afford because of the substantial revenues or operational efficiencies that accrue when large global demands are satisfied.
However, improvements in the infrastructure needed to support such requirements have not kept pace with the demand. Growth in the hardware market is driven by growing demand for multimedia applications. Demand for multimedia applications is the result of a convergence of expanded processing power, better software programming and the spread of telecommunications computing networks.
Telephone and cable companies face a continuing need to upgrade their switching and distribution networks in response to this high demand. Corporate and institutional local area networks and computing facilities are often overwhelmed by data because of equipment that was not designed to handle the data requirements needed to remain competitive in today's industrial and social climate. For these businesses, and soon the information economy in general, system crashes and slowdowns are likely to increase as current trends continue. The problem reaches far beyond the confines of individual, institutional, corporate, or even national boundaries.
As noted above, the use of, and need for, inexpensive, ubiquitous, and uninterrupted processing power and communications bandwidth is likely to continue into the foreseeable future. As telecommunications networks increase their throughput capacity, becoming more affordable and accessible, the evolutionary progression from stand-alone computers, to network computers, to on-line tele-computing is also likely to accelerate. However, this progression will require new solutions to improve the current infrastructure, which is perilously overburdened at every level.
One methodology that is being developed to increase processing power and bandwidth is parallel computing. Parallel computing uses multiple processors working in parallel on a single computing task. These processors can be linked together within a single computer, or they can be housed separately in a cluster of computers that are linked together in a network. The advantage of parallel computing over traditional, single-processor computing is that it can tackle problems faster and with greater power. For parallel computing to work, however, software and operating systems had to be re-developed within the context of multiple processors working together on one or more tasks. Standards have been developed which ensure that parallel computing users can achieve scalable software performance independent of the machine being used.
As technology has evolved, parallel processing has become a significant segment of the server market, and a growing segment of the desktop PC and workstation market. Sales of workstations and PCs have grown rapidly as the cost of the machines has dropped and their power and functionality have increased. Also fueling this trend has been the proliferation of graphically-oriented, scalable operating systems, such as Sun Microsystems Solaris, Unix, and Linux. Advanced parallelizing resources, such as Portland Group's Fortran and C++ compilers, provide a development environment for porting existing code into parallel scalable software, and for creating new software which maximizes the benefits of distributed processing. The overall effect of these changes has been to deliver increased computing power and flexibility directly to the end user via a desktop computer, while enabling the user to access and process large amounts of data via the cluster or network to which they are connected.
However, conventional network architectures yield communication bandwidths that make highly distributed numerical processing inefficient. Typical parallel programming environments have communications delays of several milliseconds. Fully exploiting the underlying advantages of parallel computing is a challenge that has eluded computer science and applications developers for decades. Developers have had to choose between the tightly coupled architecture and high efficiency of the supercomputer, or the flexibility, scalability, and cost performance of a cluster of PCs.
The execution of computer instructions over multiple processors in supercomputers and massively parallel processors has historically been accomplished by duplicating critical hardware such as memory and input-output (I/O) subsystems. These types of systems offer excellent performance, but are expensive. Moreover, low-volume manufacturing results in a significant cost/performance disadvantage, and engineering lag time may cause a technological gap between products finally appearing on the market and currently available microprocessors.
Networks of servers, workstations, and PCs may offer a cost-effective and scalable alternative to monolithic supercomputers. Using new operating systems and compilers, the bundling together of a cluster of desktop PCs and/or workstations into a parallel system has proven to be an effective solution for meeting the growing demand for computing power. Scalability, the ability to add additional processing nodes to a computing system, may be particularly essential for those systems involved in the delivery of World Wide Web information, due to the fact that Web traffic and the number of users is increasing dramatically. Future Web servers will have to deliver more complex data, voice, and video as subscriber expectations increase. Large scale systems are being built that consist of clusters of low cost computers that communicate with one another through a system area network (SAN). Clusters enable scalability to thousands of nodes, and can exploit the parallelism implicit in serving multiple simultaneous users or in processing large queries involving many storage devices. Thus, clusters can operate as a single system for tasks such as database and on-line transaction processing.
As compared to supercomputers and mainframes, cluster computing systems have the advantages of physical modularity, insulation from obsolescence, physical and logical scalability (expandability), physical and logical upgradability, and improved cost performance. However, cluster computing systems generally have less communication bandwidth, more contingencies and bottlenecks in the network protocol, many redundant and unused components, and a larger physical footprint.
SUMMARY OF THE DISCLOSURE
Therefore, it is an advantage of embodiments of the present invention to provide a system and method for cartridge-based, geometry-variant, scalable electronic systems that have the modularity, flexibility, upgradability, and cost performance of a scaleable cluster array, while yielding the physical compactness, inter-processor communications, and extended computational
3C Interactive, Inc.
Foley & Lardner
Talbott David L.
Vigushin John B.
LandOfFree
System and method for cartridge-based, geometry-variant... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System and method for cartridge-based, geometry-variant..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method for cartridge-based, geometry-variant... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2998322