Microsoft Technology Licensing, LLC

Adam B Glass

Eric Clarence Peterson

Austin N Donnelly

Antony Ian Taylor Rowstron

Stephen Francis Heil

Sergey Legtchenko

David Alex Butler

John R Douceur

Shobana Balakrishnan

David T Harper

Gregory S Miller

Jacob Kappeler Oshins

Donald M Gray

Craig M Dowell

David Tennyson Harper, Iii

George H Shaw

Robert B Nelson

Josh Oshins

James L Green

Shobana M Balakrsihnan

Shobana M Balakrishnan

Austin Nicholas Donnelly

David T Iii Harper

Low cost storage for write once read rarely data is described. In an embodiment a storage device comprises a plurality of hard disk drives connected to a server via an interconnect fabric. The storage device comprises a cooling system which is only capable of cooling a first subset of the hard disk drives and a power supply system which is only capable of powering a second subset of the hard disk drives and in some examples, the interconnect fabric may be only capable of providing full bandwidth for a third subset of the hard disk drives. Each subset may comprise only a small fraction of hard disk drives. A control mechanism, which may be implemented in software, is provided which controls which hard disk drives are active at any time in order that the constraints set by the cooling and power supply systems and interconnect fabric are not violated.

Low cost storage for rarely read data

Various technologies and techniques are disclosed for providing and interacting with a centralized wake service. A server-side wake service is provided that is operable to allow applications to subscribe to the wake service. The wake service receives a wake request directed to a particular computing device from a particular one of the applications. The wake request is forwarded from the wake service to a client-side communication service on the particular computing device if forwarding is determined to be appropriate. When the client-side communication service on the particular computing device receives the wake request while in a reduced power mode, the computing device wakes up and an appropriate response is determined.

Centralized service for awakening a computing device

A system and method for an ejection failure mechanism is provided. The system receives a request to eject an ejectable resource, and, provides information associated with a failure of the ejection of the ejectable resource, if ejection of the ejectable resource is unsuccessful. The system thus provides a deterministic mechanism through which information associated with failure of the ejection of an ejectable resource can be communicated. As such, an initiator of the request to eject can receive information associated with a cause of the ejection failure.

Ejection failure mechanism

The present invention utilizes a topology application programming interface (API) to provide relation information describing processor and platform topology to an executable program via a standardized interface. By providing topology information, the topology API allows the program to optimize its performance based upon the information without having to be explicitly aware of the actual platform architecture. The present invention also provides an expandable topology API that allows for future expansion of information type, without altering the standardized interface. In another instance of the present invention, the topology API collects, processes, and provides topology information about an underlying hardware architecture.

Systems, methods, and apparatus for indicating processor hierarchical topology

In a network of interconnected multimedia source, transfer, and sink ports, the described subject matter defers port parameter selection until substantially all relevant information is available. Specifically, parameters are negotiated between multiple ports by specifying parameter sets corresponding to the ports. Each parameter set is expressed as a list of constraints on allowable values or ranges of values. When connecting multiple ports, the constraints on their parameter sets are conjoined to form a parameter set intersection. The parameter sets of the ports are limited to values included in the parameter set intersection. Parameter sets relating to other ports are potentially limited based on the intersection results. The conjoining and limiting operations are repeated until the various parameter sets exhibit no further limitations in response to further repetitions. This process is repeated for all port interconnections. Subsequently, values for the port parameters are selected from the limited parameter sets.

Negotiating optimum parameters in a system of interconnected components

Components in a system have interconnectable ports, groups of which are to be interconnected. A method of negotiating negotiable parameters between a plurality of the ports includes specifying parameter sets corresponding respectively to the ports. Each parameter set is expressed as a list of constraints on allowable values or ranges of values. When connecting a group of ports, the constraints on their parameter sets are conjoined to form a parameter set intersection. The parameter sets of the ports are then limited to those values included in the parameter set intersection. In addition, parameter sets relating to other ports are potentially limited based on the results of the intersection. The conjoining and limiting steps are repeated until the various parameter sets exhibit no further limitations in response to further repetitions, and this process is repeated for all port interconnections. Subsequently, values for the negotiable parameters are selected from the limited parameter sets.

A method is described for synchronously rendering digitized media streams. Each digitized media stream is made up of a sequence of media samples having media-specified timing. The described method includes calculating presentation times for media samples of different media streams based in part on the media-specified timing of the media samples and also based in part upon the desired synchronization of the different media streams relative to each other. The calculated presentation times indicate when the media samples should be rendered relative to a common clock reference. The method further includes attaching a media sample's calculated presentation time to the media sample, and then routing the media sample to a sink component for rendering. The sink component renders the respective media samples of the digitized media streams at the approximate presentation times of the samples relative to the common clock reference.

Multi-media synchronization

Growing community of inventors

Woodinville, WA, United States of America

Adam B Glass