Galactic Computing Corporation Bvi/bc

Shanghai Redneurons Co., Ltd.

Galactic Computing Corporation

Galactic Computing Corporation Bvi/ibc

Yuefan Deng

Stephen J Engel

Alexander Korobka

Philip S Smith

Kitrick B Sheets

Joseph Guistozzi

Thomas O'Brien

Peng Zhang

Joseph Giustozzi

The present invention provides a mixed torus and hypercube multi-rank tensor expansion method which can be applied to the communication subsystem of a parallel processing system. The said expansion method is based on the conventional torus and hypercube topologies. A mixed torus and hypercube multi-rank tensor expansion interconnection network is built up by means of supernodes equipped with expansion interfaces. This method not only provides more bisection bandwidth to the entire system but also improves the long-range communication and global operations. Affirmatively, this expansion method can achieve better scalability and flexibility for the parallel system for a given system size.

Mixed torus and hypercube multi-rank tensor expansion method

A hosted service provider for the Internet is operated so as to provide dynamic management of hosted services across disparate customer accounts and/or geographically distinct sites.

Method and system for providing dynamic hosted service management across disparate accounts/sites

A system for distributing incoming client requests across multiple servers in a networked client-server computer environment processes all requests as a set that occur within a given time interval and collects information on the attributes of the requests and the resource capability of the servers to dynamically allocate requests in a set to the appropriate servers upon completion of the time interval. Preferably, a request table collects at least two requests incoming within a predetermined time interval, a request examiner routine analyzes each collected request with respect to at least one attribute, a system status monitor collects resource capability information of each server in a resource table and an optimization and allocation process distributes collected requests in the request table across the multiple servers upon completion of said time interval based on an optimization of potential pairings of the requests in the request table with servers in the resource table.

System for balance distribution of requests across multiple servers using dynamic metrics

The invention provides an ultra-scalable supercomputer based on MPU architecture in achieving the well-balanced performance of hundreds of TFLOPS or PFLOPS range in applications. The supercomputer system design includes the interconnect topology and its corresponding routing strategies, the communication subsystem design and implementation, the software and hardware schematic implementations. The supercomputer comprises a plurality of processing nodes powering the parallel processing and Axon nodes connecting computing nodes while implementing the external interconnections. The interconnect topology can be based on MPU architecture and the communication routing logic as required by switching logics is implemented in the FPGA chips while some modular designs for accelerating particular traffic patterns from applications and meliorating the communication overhead are able to be deployed as well.

Ultra-scalable supercomputer based on MPU architecture

A system for distributing incoming client requests across multiple servers in a networked client-server computer environment processes all requests as a set that occur within a given time interval and collects information on both the attributes of the requests and the resource capability of the servers to dynamically allocate the requests in a set to the appropriate servers upon the completion of the time interval. Preferably, the system includes a request table to collect at least two requests incoming within a predetermined time interval. A request examiner routine analyzes each collected request with respect to at least one attribute. A system status monitor collects resource capability information of each server in a resource table. An optimization and allocation process distributes collected requests in the request table across the multiple servers upon completion of said time interval based on an optimization of potential pairings of the requests in the request table with the servers in the resource table. The optimization and allocation process preferably analyzes metrics maintained in the request table and resource table as part of a relational database to allocate requests to servers based on a minimization of the metric distance between pairings of requests and servers. Preferably, the request table is part of a dynamic, relational database and a process of statistical inference for ascertaining expected demand patterns involving said the attributes adds predictive information about client requests as part of the request examiner routine.

A hosted service provider for the Internet is operated so as to provide dynamic management of hosted services across disparate customer accounts and/or geographically distinct sites. For each of a plurality of customer accounts, a plurality of individual servers are allocated to a common administrative group defined for that customer account. Each administrative group is configured to access software and data unique to that customer account to provide hosted services for that customer account. The system automatically monitors the performance and health of the servers in each administrative group. At least one server from a first administrative group is automatically and dynamically reallocated to a second administrative group in response to the automatic monitoring. The automatic and dynamic reallocation of servers is accomplished by setting initialization pointers for the reallocated servers to access software and data unique to the customer account for the second administrative group, and then reinitializing the reallocated servers such that the reallocated servers join the second administrative group when restarted. Preferably, the performance and health of the servers in each administrative group are monitored over a separate out-of-band communication channel dedicated to interconnecting the servers as an administrative group. Each administrative group includes a local decision software program that communicates with a master decision software program that determines when and how to dynamically reallocate servers to different administrative work groups in response to usage demands, available resources and service level agreements for different customer accounts.

Growing community of inventors

Setauket, NY, United States of America

Yuefan Deng