Amazon Technologies, Inc.

Other

Intel Corporation

Michael Jon Moen

Brandyn David Giroux

Darin Lee Frink

Christopher Strickland Beall

Ligong Wang

Max Jesse Wishman

Kevin M Bailey

Anthony Nicholas Liguori

Robert Charles Swanson

David Edward Bryan

Felipe Enrique Ortega Gutierrez

Jason Alexander Harland

Christopher Nathan Watson

Daniel J Ragland

Roey Rivnay

Christopher James BeSerra

Gavin Akira Ebisuzaki

Vijay Patel

Jing Wang

Afshin Majd

Robert Hastings

Asmae Mhassni

Ravi Akundi Murty

Chen An

Yaotsung Lee

Edward R Hudson

Andre L Nash

An interposer card and a virtualization offloading card are provided for installation in a third-party server to integrate the third-party server into a cloud service provider network. The interposer card includes a baseboard management controller that interfaces with a management console of the cloud service provider network. This allows the third-party server to be converted into a server controlled by the cloud service provider network. Additionally, the baseboard management controller of the interposer card acts as a firewall between the third-party server and a management control network of the cloud service provider network. The interposer card and the virtualization offloading card are installed in a chassis of the third-party server via an expansion slot without requiring modification of the hardware or firmware of the third-party server.

Hardware for integration of servers into a management network

A computer system includes symmetrical sets of motherboard serial channels which couple processor devices on a motherboard with a common serial link interface. The common serial link interface can be coupled with an endpoint device to establish symmetrical serial links between the endpoint device and the processor devices. The computer system can include a riser card which can be coupled with the serial link interface. The riser card can include an endpoint device interface and serial channels which can couple the processor devices with the endpoint device via symmetrical limited selections of the motherboard serial channels. The riser can include additional interfaces which can couple the processor devices with additional expansion devices.

Symmetrical multi-processor serial links

A power supply system can include a plurality of battery backup units and a control unit. The battery backup units may be utilized to supply backup energy to a power supply unit, for example, in case of a power outage event. The control unit can gauge a respective capacity, temperature, age, and/or other characteristic of each respective battery backup unit of the plurality of battery backup units. The control unit can designate a particular battery backup unit of the plurality of battery backup units as a spared unit, e.g., to be spared in the power outage event. The designation may be based at least in part on the respective capacity or other characteristic of each respective battery backup unit gauged.

Extending battery backup unit system service life

A rules-based mechanism is described for powering down racks in an ordered and autonomous way in a data center. Power shelf controllers (PSCs), on different racks or on the same rack, communicate together through a network, called the PSC network, separate from the data network. The PSCs are aware of the other PSCs that share the same input power domain. When the racks are configured for use, each PSC is assigned a priority value, based upon the management provisioning layer assignment. Each PSC creates a table of all the other PSCs and tracks each assigned priority value. When a power event occurs, the PSC can power down components within the rack in accordance with the priority table. Recovery can also be carried out in conformance with the priority table.

Power management network for communication between racks in a data center

Systems and methods for managing a collection of power supply units (PSUs), such as those used in a datacenter environment, can include detecting changes in an output load on the PSUs to less than a target output load that is based on a target efficiency of the PSUs. In response to the indication that the output load is decreasing, PSUs can be selectively deactivated so that that an average output load of remaining, active PSUs of the collection of PSUs remains above the target output load corresponding to the target efficiency of the PSUs.

Power supply shedding for power efficiency optimization

It may be beneficial in the case of an undesired lapse of external power to provide a back-up power supply to protect electronic components in an electronic equipment rack. This can frequently be difficult and/or costly due to the necessary addition of electronic infrastructure, such as cabling and/or logic. One means of overcoming this obstacle is to utilize unused or reserved conductors in an already utilized management cable to convey a loss-of-power signal to the components in tandem with the existing cable signals.

In-rack network for power signals

Retention mechanisms for retaining compute components within server racks are described. In one example, a retention mechanism includes a latching component configured to move between a locked position and an unlocked position. The retention mechanism also includes a retaining component configured to position the latching component in the unlocked position. Movement of the retaining component releases the latching component thereby allowing the latching component to move into the locked position.

Rack component retention mechanisms

An assembly for containing and detecting leaks from a coolant line includes a basin with side and end walls, and at least one U-shaped cavity in an end wall for receiving a coolant line of a liquid cooling system. A bottom of the basin includes at least one sloped interior surface arranged to divert water in the basin to a local low point, where a sensor is positioned and operable to detect liquid in the collection basin. A leak detection system can employ a leak detection assembly to detect and/or locate a leak.

Leak detection and containment for liquid cooled computing

Latching components for retaining rack component cases within server racks are described. In one example, a latching component is installed at an interior surface of a rack component case. The latching component includes a protrusion. The latching component is biased in a locked position in which the protrusion is outside the rack component case. The latching component is moveable from the locked position to an unlocked in which the protrusion is removed from outside the rack component case using an unlatching component.

Rack component latches

A system includes hardware logic circuit configured to execute power control software. The hardware logic circuit executes the power control software to receive a request from a processing unit for delivery of a power level corresponding to a first power state and determine that the processing unit should operate at a second power state, where the second power state is different than the requested first power state. The hardware logic circuit also executes the power control software to cause a power supply to deliver a power level corresponding to the second power state to the processing unit.

Power delivery for processing units

A method for migrating a power shelf between being supplied single-phase power and three-phase power includes coupling a second power adapter to an input of a power shelf of a rack while power is being supplied to the power shelf via a first power adapter coupled to another input of the power shelf. The first power adapter includes a multi-pin connector coupled to the input and three other connectors each coupled to respective single-phase power sources and the second power adapter includes a multi-pin connector configured to couple with the other input and another connector configured to couple with a three-phase power source.

Rack power shelf migration between single-phase and three-phase power

A connection panel for enabling quick change over of a network device comprises a body and an engaging device. An exterior side of the body has multiple sockets, each of which can receive a plug connector. The panel is configured for positioning at one end of a rack space for the network device. The panel's interior side has first network device connections configured to mate with second network device connections on the network device. The engaging device has at least one member extending between the network device and the panel, and is repositionable between at least an engaged position in which the network device is moved towards and mechanically coupled to the panel with the second network device connections urged into engagement with and electrically connected to the first connections, and a disengaged position in which the network device is disengaged from the panel and removable from the rack space.

Connection panel and methods for decreasing service time for server rack network devices

Systems and methods consistent with the present disclosure include techniques for dynamic system performance tuning (DSPT). Techniques for DSPT include identifying an active software application during a user session and applying an application-specific profile that defines different system-hardware operating states of a computing system to enhance the performance of the active software application.

Growing community of inventors

Olympia, WA, United States of America

Michael Jon Moen