Hewlett-packard Development Company, L.p.

Hewlett Packard Enterprise Development LP

Antonio Lain

Patrick Goldsack

Nigel John Edwards

Adrian John Baldwin

Christopher Tofts

Richard Taylor

Paolo Faraboschi

Moray McLaren

Zhikui Wang

Sai Rahul Chalamalasetti

Jose Renato G Santos

Brian Quentin Monahan

Francisco Guerrero

Alberto Such

Joseph A Fisher

Anshuman Goswami

Viacheslav Borisov

A method for allocating resources includes determining that an initial allocation of memory bandwidth for one or more computing jobs fails a performance metric. The memory bandwidth provides access to a global memory pool for multiple legacy processors across a memory fabric. The method also includes determining a new allocation of memory bandwidth for the computing jobs that meets the performance metric. Additionally, the method includes assigning the new allocation of memory bandwidth to the computing jobs. The method further includes executing the computing jobs using the new allocation of memory bandwidth.

Allocating resources of a memory fabric

Systems and methods associated with neural network verification are disclosed. One example method may be embodied on a non-transitory computer-readable medium storing computer-executable instructions. The instructions, when executed by a computer, may cause the computer to train a neural network with a training data set to perform a predefined task. The instructions may also cause the computer to train the neural network with a sentinel data set. The sentinel data set may cause the neural network to provide an identification signal in response to a predefined query set. The instructions may also cause the computer to verify whether a suspicious service operates an unauthorized copy of the neural network. The suspicious service may be verified by extracting the identification signal from responses the suspicious service provides to the predefined query set.

Neural network verification

A connection policy for a communications network has a local connection policy indicating which paths between a given one of the nodes (computer A, router A, host) and others of the nodes (computers B, C, filters B, B, C, C, hosts) are allowable paths, by a symbolic expression of ranges endpoint addresses and other local connection policies in respect of other nodes. It is implemented in a distributed manner by determining, for the given node, which of the allowable paths, are dual authorized as allowable by the other local connection policy relating to the other node at the other end of that path, by Boolean operations on the symbolic expressions. For a given message for a given path between two of the nodes having their own local connection policies, both of these nodes determine whether the given path is currently dual authorized. This can provide reassurance that changes in versions of the connection policy won't transiently open a risk of undetected unwanted communication.

Distributed network connection policy management

Illustrated is a system and method to receive an instruction to access a set of identified data structures, where each identified data structure is associated with a version data structure that includes annotations of particular dependencies amongst at least two members of the set of identical data structures. The system and method further comprising determining, based upon the dependencies, that a version mismatch exists between the at least two members of the set of identified data structures, the dependencies used to identify a most recent version and a locally cached version of the at least two members. The system and method further comprising delaying execution of the instruction until the version mismatch between the at least two members of the set of identified data structures is resolved through an upgrade of a version of one of the at least two members of the set of identified data structures.

Version mismatch delay and update for a distributed system

A system has a virtual overlay infrastructure mapped onto physical resources for processing, storage and network communications, the virtual infrastructure having virtual entities for processing, storage and network communications. Virtual infrastructures of different users share physical resources but are isolated. Each infrastructure has its own infrastructure controller to create and configure the infrastructure. It has a user accessible part (CFC) for configuration of that user's infrastructure, and a user inaccessible part (UFC) able to access the mapping and the physical resources. This increases user control to ease system administration, while maintaining security by limiting access to the mapping.

Controlling virtual overlay infrastructure

In a method for managing a memory segment through use of a memory virtual appliance, data is encapsulated with the memory virtual appliance, in which the memory virtual appliance comprises a virtual machine configured to manage a memory segment in a physical memory. In addition, the memory virtual appliance is implemented using a virtualization wrapper comprising computer readable code enabling the encapsulated data to be shared among a plurality of clients. Moreover, the encapsulated data is stored in the memory segment controlled by the memory virtual appliance.

Managing a memory segment using a memory virtual appliance

A system has a virtual overlay infrastructure mapped onto physical resources for processing, storage and network communications, the virtual infrastructure having virtual entities for processing, storage and network communications. Each virtual infrastructure can be passivated by suspending applications, stopping operating systems, and storing state, to enable later reactivation. This is simpler for a complete virtual infrastructure than for groups of virtual entities and physical entities. It enables cloned virtual infrastructure to be created for testing, upgrading or sharing without risk to the parent. On failure, reversion to a previous working clone is feasible.

Managing virtual overlay infrastructures

A method and apparatus is provided for consolidating cryptographic key updates, the consolidated update information enabling, for example, a returning member of a secure group who has been offline, to recover the current group key, at least in most cases. The unconsolidated key updates each comprise an encrypted key, corresponding to a node of a key hierarchy, that has been encrypted using a key which is a descendant of that node. The key updates are used to maintain a key tree with nodes in this tree corresponding to nodes in the key hierarchy. Each node of the key tree is used to store, for each encrypting key used in respect of the encrypted key associated with the node, the most up-to-date version of the encrypted key with any earlier versions being discarded. The key tree, or a subset of the tree, is then provided to group members.

Cryptographic key update management method and apparatus

A method of authorizing a computing entity comprises obtaining at the authorizing entity a delegation chain of intermediate elements through which an authorisable entity asserts authorization eligibility in the form of a sequence of locally assigned labels for the elements. The method further comparing the label sequence against a label sequence template.

Method of authorising a computing entity

A computer implemented process for identifying conflicts between policies for a method invocable by an object is described, in which the method is a member of at least one statically defined method grouping having a corresponding policy. The process comprises, for each method grouping, obtaining each method which is a member of the method grouping, associating therewith the corresponding method grouping policy, and comparing policies associated with methods which are members of two or more method groupings to identify conflicts.

Computer implemented process for identifying conflicts between policies

Security keys for the provision of a secure service such as content provision are generated in an ancestral hierarchy, so that invalidation of a key in the hierarchy results in a need to reconfigure all other keys in the hierarchy to the extent they share common ancestry. When a user subscription to the service lapses, a decision on invalidation of their key is based in a determination of whether it's more costly to the subscriber to invalidate the key, or continue providing an unpaid-for service. Keys can be allocated to users from domains of the hierarchy on the basis of their economic value to the provider, with higher value users being allocated keys from domains which share fewer common ancestors with other users of other domains than those users share with each other, to minimise inconvenience to high value users of key reconfiguration.

Management of security key distribution

Security keys for the provision of a secure service such as content provision are generated in an ancestral hierarchy, so that invalidation of a key in the hierarchy results in a need to reconfigure all other keys in the hierarchy to the extent they share common ancestry. When a user subscription to the service lapses, a decision on invalidation of their key is based in a determination of whether it's more costly to the subscriber to invalidate the key, or continue providing an unpaid-for service. Keys can be allocated to users from domains of the hierarchy on the basis of their economic value to the provider, with higher value users being allocated keys from domains which share fewer common ancestors with other users of other domains than those users share with each other, to minimize inconvenience to high value users of key reconfiguration.

The present disclosure relates to a system and method for isolating applications from each other. In one arrangement, the system and method pertain to intercepting application instructions, determining if an application instruction calls for an illegal action, and preventing the illegal action from occurring if the application instruction calls for the illegal action. In some arrangements, the system and method further pertain to determining if the application instruction contains an indirect memory reference and, if so, modifying the application instruction to enable the application instruction to determine itself whether an illegal operation is to be performed.

System and method for isolating applications from each other

While powerful in diffusion at one resolution to print photos at a finer resolution, the invention is not thus limited. It defines superpixels ('spels') for each desired colorimetric level, generates/receives image data, renders by finding levels for image positions, and prints an image using selected spels. One invention aspect finds a randomized value at each found level and uses the value to select the spel from plural ones for each level. Another aspect derives/maintains a randomized-value matrix; and maps a matrix location to an image position, to select a random value at that location and spel for that position. Another uses the value in common for all planes to select a spel for each plane at the found level&#x2014;compatible spels for different planes, to coordinate color placement in planes. Another controls defining/selecting for a blue-noise property of spels in aggregate. In another, spels defined for a level vary in value to yield nonintegral color quanta. Preferences: Rendering is 1D per color, plus a dummy dimension holding the value (a least-significant bit from rendering, in a color dimension but decorrelated from levels)&#x2014;and derives/maintains the matrix, derived/corrected for blue noise and including small interleaved 1D matrices tiled across and wrapped around the larger. Mapping uses values as pointers into dimensions of a spel table, and color-plane identification as a pointer into a third table dimension; it uses a common value in all planes to avoid drop-on-drop. Spels are Fourier-screened.

Growing community of inventors

Bristol, United Kingdom

Antonio Lain