Google Inc.

Phoenix Technologies Ltd.

Advanced Green Computing Machines-ip, Limited

Eric Christopher Anderson

Pawel Opalinski

John Felton

Brandon Player Iles

David H Nguyen

Tung M Nguyen

Syed Najaf Rizvi

Henri Hayim Farhi

Adam Drobot

David Pitard

A multi-node computer system, comprising: a plurality of nodes, a system control unit and a carrier board. Each node of the plurality of nodes comprises a processor and a memory. The system control unit is responsible for: power management, cooling, workload provisioning, native storage servicing, and I/O. The carrier board comprises a system fabric and a plurality of electrical connections. The electrical connections provide the plurality of nodes with power, management controls, system connectivity between the system control unit and the plurality of nodes, and an external network connection to a user infrastructure. The system control unit and the carrier board provide integrated, shared resources for the plurality of nodes. The multi-node computer system is provided in a single enclosure.

High density multi node computer with improved efficiency, thermal control, and compute performance

A method for providing metadata to a search engine for a document that is not in a mark-up language receives a request for contents of the document and locates metadata associated with the document. The method further creates name-value pairs for the metadata and provides to the search engine server a response comprising the name-value pair in an HTTP (or HTTPS) header and the contents of the document. In other implementations, a method includes sending a request for contents of the document and receiving a response to the request comprising an HTTP header with metadata about the document in a name-value pair and the document's content. The method also includes extracting the name-value pair from the HTTP header, creating a mark-up language tag for the name-value pair, and providing the make-up language tag and the contents of the document in a mark-up language format to a search index creation component.

Methods and systems for providing custom crawl-time metadata

A method for providing metadata to a search engine for a document that is not in a mark-up language includes sending a request for data about the document and receiving a response to the request that has a Hyper-Text Transfer Protocol (HTTP or HTTPS) header that includes metadata associated with the document in a name-value pair and the document's content. The method also includes extracting the name-value pair from the HTTP-header and creating a mark-up language tag for the name-value pair and providing the make-up language tag and the contents of the document in a mark-up language format to a search index creation component.

A system for searching files stored in a closed file source that is not accessible via a web crawler obtains file identifiers for files stored in the file source and creates a unique URL for each of the identifiers. Each URL may be based on a file identifier and a domain portion of a URL associated with the system. The system may provide the unique URLs to a search engine. The system may respond to a crawl request from the search engine for a particular URL by converting the URL back into a file identifier, obtaining the contents of the file, creating an HTTP response from the contents of the file, and returning the response to the search engine. The system may respond to a request for a seed URL with a plurality of URLs as links in a single HTTP response.

Adapting content repositories for crawling and serving

A method including copying a first application from a first non-volatile memory to a second non-volatile memory and setting the first application to be a start-up application. The method further includes booting-up an operating system and executing the first application. The method also includes determining if a suitable connection exists, determining if a needed bandwidth of the suitable connection is available, and downloading a second application if the suitable connection exists and the needed bandwidth of the suitable connection is available. The method then includes executing the second application. An apparatus for performing the method is also disclosed.

Method and apparatus for automatically installing and configuring software on a computer

The present invention is a method and apparatus to determine loading time of an operating system in a computer system. A basic input and output system (BIOS) retrieves a first time value from a real-time clock. The BIOS loads the OS into memory of the computer system. The OS has a device driver. The device driver retrieves a second time value from the real-time clock. A time difference is computed based on the first and second time value to obtain the loading time of the OS.

Determining loading time of an operating system

A method for managing power in an electronic system having a plurality of input/output devices, each of the input/output devices having a full power-on state and at least one power reduction state and each being controlled by an associated device driver, the method comprising the steps of: initializing a power management logic separate from the device drivers to receive at least one time-out value for each of the device drivers to be subject to power management; assigning a different timer in the power management logic to a different one of the device drivers to be subject to power management, the timers being disposed external of the device drivers; initializing each of a plurality of the assigned timers to individual predetermined value set in accordance with the at least one time out value of the assigned device driver for that timer; changing the individual predetermined value held in each of the timers at a predetermined interval; monitoring a plurality of the timers and determining when the predetermined value in one of the timers indicates that the time-out value for the assigned device driver has elapsed, and sending a power reduction instruction to the assigned device driver; and resetting a given one of the timers to its individual predetermined value at any time when the device driver associated with the timer indicates a usage event of its associated device driver, and powering the associated device to a higher power level if it is not at full power-on.

Growing community of inventors

Sunnyvale, CA, United States of America

Eric Christopher Anderson