Abstract: Computing system enhancements make container instantiation faster, reduce layer content storage demands, and make more container image formats available. A container instantiation location sends a container image pull request to a container registry, receives an image manifest, sends a layer mount request to the registry instead of a layer content download request, receives a layer mount, optionally repeats for additional layers, creates a union file system spanning the layers, and launches a container process based on the union file system without first downloading all the layer content.

Abstract: A computer-implemented method includes: initializing model parameters for training a neural network; performing a forward pass and backpropagation for a first minibatch of training data; determining a new weight value for each of a plurality of nodes of the neural network using a gradient descent of the first minibatch; for each determined new weight value, determining whether to update a running mean corresponding to a weight of a particular node; based on a determination to update the running mean, calculating a new mean weight value for the particular node using the determined new weight value; updating the weight parameters for all nodes based on the calculated new mean weight values corresponding to each node; assigning the running mean as the weight for the particular node when training on the first minibatch is completed; and reinitializing running means for all nodes at a start of training a second minibatch.

Abstract: Computing system enhancements make container instantiation faster, reduce layer content storage demands, and make more container image formats available. A container instantiation location sends a container image pull request to a container registry, receives an image manifest, sends a layer mount request to the registry instead of a layer content download request, receives a layer mount, optionally repeats for additional layers, creates a union file system spanning the layers, and launches a container process based on the union file system without first downloading all the layer content.

Abstract: Consumption of power by device components is profiled on a per-session basis of user interaction. A session engine determines start of a user interaction session with the device (e.g., by detecting screen turn on). The engine generates a session identifier (SID) and broadcasts same to device component(s). In response to the SID, the component(s) record power consumption data. Upon receiving a signal indicating end of the user interaction session (e.g., screen turn off), the session engine broadcasts a notice allowing component(s) to stop recording power consumption data. The components communicate aggregated data to the session engine for storage in a centralized location (e.g., trace buffer). In response to a query posed within a command line prompt, stored data may be parsed on a per-session basis for inclusion in a detailed report of power consumption. Power consumption inefficiencies endemic to components and/or user behaviors may be thus be identified and remediated.

Abstract: Consumption of power by device components is profiled on a per-session basis of user interaction. A session engine determines start of a user interaction session with the device (e.g., by detecting screen turn on). The engine generates a session identifier (SID) and broadcasts same to device component(s). In response to the SID, the component(s) record power consumption data. Upon receiving a signal indicating end of the user interaction session (e.g., screen turn off), the session engine broadcasts a notice allowing component(s) to stop recording power consumption data. The components communicate aggregated data to the session engine for storage in a centralized location (e.g., trace buffer). In response to a query posed within a command line prompt, stored data may be parsed on a per-session basis for inclusion in a detailed report of power consumption. Power consumption inefficiencies endemic to components and/or user behaviors may be thus be identified and remediated.