Abstract: A first representation of a feature processing recipe is received at a machine learning service. The recipe includes a section in which groups of variables on which common transformations are to be applied are defined, and a section in which a set of transformation operations are specified. The first representation of the recipe is validated based at least in part on a library of function definitions supported by the service, and an executable version of the recipe is generated. In response to a determination that the recipe is to be executed on a particular data set, a set of provider network resources is used to implement a transformation operation indicated in the recipe.

Abstract: Features are disclosed for improving the robustness of a neural network by using multiple (e.g., two or more) feature streams, combing data from the feature streams, and comparing the combined data to data from a subset of the feature streams (e.g., comparing values from the combined feature stream to values from one of the component feature streams of the combined feature stream). The neural network can include a component or layer that selects the data with the highest value, which can suppress or exclude some or all corrupted data from the combined feature stream. Subsequent layers of the neural network can restrict connections from the combined feature stream to a component feature stream to reduce the possibility that a corrupted combined feature stream will corrupt the component feature stream.

Abstract: An electrowetting element comprising a first fluid comprising a dye compound having the general formula:

Abstract: Methods, apparatus, and computer-accessible storage media for remotely updating an executing process that receives I/O requests on I/O port(s), stores write data to a write log on local storage, and uploads data from the write log to remote storage. An update for the process is detected and downloaded, and an updated process is instantiated from the update. The current process is directed to perform a shutdown for update during an update window. In response, the current process saves its current configuration, flushes an in-memory portion of the write log to local storage, and releases its I/O port(s). The updated process loads the saved configuration, detects that the port(s) have been released, and starts accepting I/O requests on the ports. During flushing, the current process flushes current data in memory while continuing to append new write data, stops accepting new write requests, and then flushes the new write data.

Abstract: An electronics adapter and method are disclosed herein. The electronics adapter can include a plurality of interface ports, with each interface port from the device coupled to a processor from a plurality of processors, and a controller communicatively coupled to the interface ports. The controller may be configured to determine a function or transaction attributes, which are serviced by instructions executed by one of the processors. The controller may be further configured to determine at least one interface port on the adapter to transmit the transaction based on the function or the attributes using an updatable mapping between the function or the attributes and the interface port, and transmit a request for the transaction using the interface port for processing of the transaction by the processor coupled to the interface port.

Abstract: Network hardware devices organized in a Wireless mesh network (WMN) in which the network hardware devices cooperate in distribution of content files to client consumption devices in an environment of limited connectivity to broadband Internet infrastructure are described. One mesh network device includes a micro controller, RF radios, and antenna switching circuitry. The antenna switching circuitry, in response to the control signals from a micro controller, selectively couples individual ones of a first set of antennas to individual channels of the set of RF radios to communicate content data with client consumption devices and selectively couples individual ones of a second set of antennas to other individual channels of the set of RF radios to communicate content data with other mesh network devices in a WMN.

Abstract: Methods and systems for implementing a virtual device farm are disclosed. A client device is registered to participate in software testing. One or more applications are selected based on the configuration of the client device. The one or more selected applications are installed on the client device. Programmatic testing of the one or more selected applications is performed using the client device. Results of the programmatic testing are collected.

Abstract: A system, method, and computer-readable medium for point of presence (POP) based traffic surge detection and mitigation are provided. The system detects a traffic surge for a target group of resources directed at a source POP based on the target group's rank shifts and volume changes among recent time intervals. The system mitigates the detected traffic surge by identifying destination POPs with spare capacity and routing at least a portion of incoming requests for the target group of resources to the destination POPs in accordance with their spare capacities.

Abstract: Disclosed are various embodiments for monitoring, maintaining, tracking, and/or integrating inventory items in one or more systems. An inventory integration system may monitor the states, dependencies, and/or locations of inventory items across one or more data centers and/or similar structures. Issues detected may be escalated and remedial actions to the issues may be generated. The inventory integration system may communicate with various external resources in order to convey various information associated with the items of inventory and/or the data centers.

Abstract: Described are techniques for prioritizing devices to receive updates to an application executing on a device. Event data indicative of use of the application, the device, and other information associated with a user is obtained. For example, event data for a user account across many different devices may be obtained. Deployment parameters may specify how an update is to be distributed, such as maximum number of devices to deploy to, timing of deployment, and so forth. Based on the event data and the deployment parameters, a subset of the devices executing the application are selected. The update may then be distributed to the subset.

Abstract: A data center row infrastructure module enables rapid deployment of a computing room enclosure for computer systems that provides intake air circulation to, and exhaust air removal from, the computer systems in the enclosure. The data center row infrastructure module includes free-standing exhaust plenum structures encompassing opposite sides of the enclosure and a plenum module spanning a top end of the enclosure between the free-standing exhaust plenum structures. Each free-standing exhaust plenum structure includes an internal exhaust air plenum that directs exhaust air received from the enclosure through an exhaust air outlet on a top end of the free-standing exhaust plenum structure. The plenum module establishes a bottom end of a cooling air plenum above and separate from the enclosure and between the free-standing exhaust plenum structures. Some plenum modules can direct cooling air directly to the enclosure as intake air.

Abstract: Customers can utilize resources of a multi-tenant environment to provide one or more services available to various users. In order to simplify the process for these customers, the multi-tenant environment can include an infrastructure wherein a portion of the resources provide an authentication and/or authorization service that can be leveraged by the customer services. These resources can logically sit in front of the resources used to provide the customer services, such that a user request must pass through the authorization and authentication service before being directed to the customer service. Such resources can provide other functionality as well, such as load balancing and metering.

Abstract: Server computers may include one or more input/output (I/O) adapter devices for communicating with a network and/or direct-attached device. The I/O adapter device may implement processes to manage write requests in a general and flexible manner. The I/O adapter device may also implement processes to manage write requests in a fast an efficient—that is, low latency—manner. Low latency write requests processes may include determining that a write packet for a write request can be processed without additional assistance from a processor, once a processor has initiated a memory access request to fetch write data and also generated protocol information for transmitting the write packet. The I/O adapter device may then process and transmit the write packet through an offload pipeline, without interrupting a processor.

Abstract: A system, method and computer-readable medium for client-side cache management are provided. A client request for content is returned that includes executable code for generating a request for preload information. Based on processing the executable code, a client computing device requests preload information from a content delivery service provider. The content delivery service provider provides an identification of content based on resource requests previously served by the content delivery service provider. The client computing device processes the preload information and generates and obtains identified resources for maintenance in a client computing device memory, such as cache.

Abstract: An item-providing system supplies items to a user device for consumption at the user device via communication infrastructure. The device may correspond to a book reader device or other type of device. The item-providing system uses a content delivery module to deliver the items to the user device.

Abstract: Methods, systems, and computer-readable media for local-to-remote migration for virtualized graphics processing are disclosed. A virtual compute instance comprising a local GPU is provisioned from a provider network. The provider network comprises a plurality of computing devices configured to implement a plurality of virtual compute instances with multi-tenancy. A virtual GPU is attached to the virtual compute instance. The virtual GPU is implemented using a physical GPU, and the physical GPU is accessible to the virtual compute instance over a network. Graphics processing for the virtual compute instance is migrated from the local GPU to the virtual GPU. An application is executed using the virtual GPU on the virtual compute instance.

Abstract: Systems, methods, and media for presenting panel-based electronic documents are provided. In accordance with some embodiments, systems for presenting panel-based electronic documents are provided, the systems comprising: at least one processor programmed to: receive an electronic document, a definition of a first panel on a page in the electronic document, and a definition of a second panel on the page in the electronic document; control a display of the first panel based on the definition of the first panel; and transition from the display of the first panel to a display of the second panel by re-scaling the display and panning from the first panel to the second panel.

Abstract: A route may be created and provided to a client. A stopping point along the route is received, and a search is performed for at least one point of interest according to the stopping point. Further, a route is created and provided to a client, and at least one stopping point associated with the route is identified based on at least one of a distance along the route, an indication of a place on a map, a category search, and a key word search.

Abstract: A group of computers is configured to implement a block storage service. The block storage service includes a block-level storage for storing data from a set of distinct computing instances for a set of distinct users. An interface is configured to allow the set of distinct users to specify respective destinations for storing backup copies of respective data stored in the block-level storage for the distinct users. At least some of the respective destinations are for different storage systems remote from one another. A backup copy function is provided for creating backup copies of data stored in the block-level storage by the set of distinct computing instances for the set of distinct users. The backup copies are stored in different destination locations specified by respective ones of the plurality of distinct users via the interface.

Abstract: A system and method for management and processing of resource requests is provided. A content delivery network service provider receives a DNS query from a client computing device. The DNS query corresponds to a requested resource from the client computing device. The content delivery network service provider associates the client computing device with a cluster of other client computing devices. Based on routing information for the cluster, the content delivery network service provider routes the DNS query. The process can further include monitoring performance data associated with the delivery of the requested resource and updating the routing information for the cluster based on the performance data for use in processing subsequent requests from client computing devices in the cluster.