Abstract: An interconnecting module configured to be mounted in a High-Performance Computing (HPC) cabinet in order to interconnect a plurality of computing units located in the HPC cabinet. The interconnecting module comprising a chassis configured to be mounted horizontally within housings defined in the computing units, a plurality of connection units spaced vertically, each connection unit being configured to be connected to a motherboard of a computing unit, the connection units being interconnected, each connection unit comprising at least a fastening member configured to cooperate with a housing of a computing unit, a vertical rack cooperating with all the fastening members and at least a main lever mounted pivotally on the chassis, the main lever being configured to move the vertical rack vertically in order to engage/disengage all the fastening members simultaneously.

Abstract: The invention relates to a computer-implemented method for identifying a radar transmitter from a set of corresponding received pulses, each pulse being associated with a respective time of arrival. The method includes determining an observed signature of the radar transmitter based on a distribution of the time gaps between consecutive times of arrival. For each transmitter class among a plurality of predetermined transmitter classes, each transmitter class being associated with at least one expected signature, the method includes calculating a proximity score between the observed signature and each expected signature associated with said transmitter class. Each expected signature is a function of an expected distribution of the time gaps between consecutive times of transmission for said transmitter class, and for a predetermined pulse loss rate. The method also includes assigning the radar transmitter to the transmitter class associated with the expected signature that provides the best proximity score.

Abstract: The invention relates to a computer-implemented surveillance method for surveilling an area of interest using a set of drones, the surveillance method including monitoring a number of operational drones among the set of drones. If a change in the number of operational drones is detected, performing a segmentation step for segmenting the area of interest into N sub-areas, N being the current number of operational drones; and performing an affectation step for affecting each operational drone to a respective sub-area for surveilling said sub-area.

Abstract: The invention relates to a method (100) for deploying a software as a service, SaaS, application, in particular an authentication as a service, IDaaS, application, said method (100) comprising the following steps: deploying (104) a load distribution module, provided for receiving the connections to said application, and distributing said connections within a group of application nodes, each executing an instance of said application and being provided to perform at least one connection to said application, and deploying (106) at least one master application node; said method (100) further comprising at least one iteration of a step (130; 152) of adjusting, by a master application node, the number of application nodes in said group based on an index, called the load index, representative of the total load of the application nodes of said group. The invention also relates to a computer program and a system implementing such a method.

Abstract: The invention relates to a method for monitoring the operation of an application, including acquiring data that has a plurality of key performance indicators related to the current operation of the application; determining, based on the acquired data, forecasted data that includes a plurality of key performance indicators related to a future operation of the application; automatically scheduling a monitoring session targeting at least one component of the application corresponding to a group of key performance indicators, the group of key performance indicators including each key performance indicator of the forecasted data having a value moving away from a corresponding predefined value range being associated with a normal operation of the application; obtaining, from a user, using a conversational agent, information related to an anomalous behavior of the application; and modifying the scheduled monitoring session based on the obtained information.

Abstract: A method for optimizing the detection of target cases in an imbalanced dataset, including generating a series of training datasets wherein the first training dataset includes an equal ratio of non-target cases and target cases and wherein the following training datasets of the series comprise a ratio of non-target to target cases that increases for each consecutive training dataset of the series. The method also includes training the machine learning model using the machine learning algorithm on each generated training datasets of the series of training datasets and recording the obtained performance score at each iteration, determining the maximum performance score among the recorded performance scores, determining the ratio of target to non-target cases for the determined maximum performance score, and training the machine learning model using the machine learning algorithm on a training dataset having the determined ratio of target to non-target cases to obtain an optimized model.

Abstract: The invention relates to a method for migrating a customer communication document from a first customer communication management platform to a second customer communication management platform. The method automates or at least semi-automates the migration of documents between customer communication management platforms by automating the generation of an intermediate document. The method automatically detects the dynamic information and the static information displayed on a set of documents, which is a crucial point to generate the intermediate documents.

Abstract: The invention relates to a method for adapting an artificial neural network (ANN) model, previously trained for person re-identification on a source domain, on a target domain. The method includes several iterations of an adaptation phase including constructing a support set, by selecting from images of the source domain similar to images in a new set of images received from the target domain. The method also includes several iterations of a training step including determining a re-id cost on the new set of images; determining a Knowledge Distillation cost, with respect to a teacher model, on a support set constructed during a previous iteration of the adaptation phase, updating the ANN model; and updating the teacher model. The invention further relates to a computer program and a device configured to carry out such a method, and to an artificial neural network trained with such a method.

Abstract: Disclosed is a server including a plurality of information processing modules so as to form a unified platform for an operating system deployed on the server, each module having an elementary software, BIOS, provided to use configuration data when the module is started up, the server also being provided to allow at least one configuration datum to be modified, and being characterized in that it further includes a management controller, BMC, provided to store configuration data in an associated memory, and to transmit, when the at least one configuration datum is modified, such data to management controllers associated with the other modules, and in that the elementary software is provided to recover the configuration data from the management controller when the module is started up.

Abstract: The invention relates to a method for carrying out a computing job for a client on a High Performance Computing (HPC) machine. The method includes selecting, in the HPC machine, the resources for carrying out the computing job, wherein the resources form a compute environment that includes one or several computing nodes. The method also includes performing the computing job in the compute environment; and before performing the computing job, isolating the compute environment from the rest of the HPC machine so that interaction with the compute environment is restricted to the client. It also relates to an HPC machine that carries out the method.

Abstract: A method and system for intrusion detection to detect malicious insider threat activities within a network of user profiles. The method includes training a Neural Network on multiple sets of user profile data for multiple user profiles and on multiple sets of activity data of the multiple user profiles of the network, such that the Neural Network is capable of predicting for future dates activities for multiple user profiles. The method includes applying the trained Neural Network on the set of further user profile data of the further user profile, predicting an activity of the further user profile based on the multiple sets of activity data by the trained Neural Network, observing activity of the further user profile, applying the trained Neural Network on the observed activity, and detecting malicious activity for the further user profile by the trained Neural Network, if the observed activity deviates from the predicted activity.

Abstract: The invention relates to an electronic circuit comprising a substrate and at least one electronic component (3, 6), each electronic component (3, 6) including at least one data port (12) for receiving data, the electronic circuit (2) further comprising, for each data port (12): a current injection terminal (8); a breakable electrical trace (10) located in the thickness of the substrate and extending between the data port (12) and the injection terminal (8), each breakable electrical trace (10) comprising a fusible segment (20) and being configured to break, by melting of the fusible segment (20), upon the injection, at the injection terminal (8), of an electric current (I) having an intensity greater than a predetermined trace breaking threshold.

Abstract: The invention relates to a method for communicating messages between a plurality of pieces of user equipment of a set of user equipment via a communication network. The method includes, when a transmitter determines that a last payload message of a series received by a receiver, for which all payload messages preceding it in the series have also been received, is different from the last payload message of the series, rebroadcasting by the sender of the payload messages of the series following the last payload message received by the receiver.

Abstract: The invention relates to a method for communicating messages between a plurality of pieces of user equipment of a set of user equipment via a communication network. The method includes, for each receiver having determined a recipient status of a series of payload messages sent, sending to the sender by the receiver, an acknowledgment message including receive information for each payload message of the series received. In the absence of receipt of an acknowledgment message by the recipient or upon receipt of an acknowledgment message that does not include receive information for each payload message of the series, and after receipt of a status message from the recipient indicating its connection to the communication network, the method includes retransmission by the sender of each payload message not received.

Abstract: The invention relates to a method for optimizing the execution parameters of a software application on an information processing platform, consisting in iteratively optimizing said parameters at each execution of said application, in which: for each execution (1) of said application, performance measurements are made (2) and stored (3), in association with the parameters used for said execution; at the start of each execution of said application, values are determined (4, 6) for a first subset of said parameters by inference (6) from the stored measurements corresponding to a subset of the executions corresponding to a second subset of said parameters.

Abstract: Method for manufacturing a heat sink for liquid cooling system of a compute blade of a supercomputer, comprising in particular a friction-stir welding step of a base plate at a machined central rib and at a machined rim.

Abstract: The invention relates to a method of predicting a trajectory of a given satellite, including training a machine learning algorithm to predict the trajectory of the given satellite from a data set of given satellite, the algorithm being encoded in a programming language; integrating the trained algorithm, on an integrated circuit, by converting the programming language into a hardware description language; and predicting the trajectory of the given satellite given by the integrated algorithm, from a data set of the given satellite. The training and integrating are performed on the ground on a computer comprising at least one processor and the predicting is performed on board the given satellite embarking the integrated circuit.

Abstract: A method for deinterleaving radar pulses, implemented by a computer. The method includes implementing a first clustering algorithm for assigning each pulse to a corresponding first class, based on the associated frequency, duration and time of arrival. The method also includes, for each first class, the method includes estimating a respective average frequency and average pulse duration, and implementing a second clustering algorithm to group the first classes into second classes based on the corresponding average frequency and average duration. For each second class, the method includes determining a distribution of the times of arrival of the associated pulses. The method also includes implementing a third clustering algorithm for grouping together the second classes into third classes, based on optimal transport distances between the corresponding determined distributions of the times of arrival.

Abstract: The invention relates to an access management method comprising, in response to a request to access a local application (23) hosted on a second network (10): if the second network (10) is connected to a first network (8): receiving the access request by a master module (4) connected to the first network (8); determining, by the master module (4), access confirmation or denial data for access to the local application; and transmitting the determined access confirmation or denial data from the master module to the local application (23) otherwise: receiving the access request by a satellite module (6) connected to the second network (10); determining, by the master module (6), access confirmation or denial data for access to the local application (23); and transmitting the determined access confirmation or denial data from the satellite module to the local application (23).

Abstract: An aspect of the invention relates to a method of managing jobs in a information system (SI) on which a plurality of jobs run, the information system (SI) comprising a plurality of computer nodes (NDi) and at least a first storage tier (NS1) associated with a first performance tier and a second storage tier (NS2) associated with a second performance tier lower than the first performance tier, each job being associated with a priority level determined from a set of parameters comprising the node or nodes (NDi) on which the job is to be executed, the method comprising a step of scheduling the jobs as a function of the priority level associated with each job; the set of parameters used for determining the priority level also comprising a first parameter relating to the storage tier to be used for the data necessary for the execution of the job in question and a second parameter relating to the position of the data necessary for the execution of the job (TAi) in question.