Abstract: Method and system for training a machine learning model based on a training dataset formed by data objects distributed across a virtual object storage service. The method comprises fetching from the virtual object storage service, the training dataset; copying the fetched training dataset on a first local storage device and maintaining a list of modifications executed on the training dataset that occurred on the virtual object storage service. The method comprises, upon receiving a request to initiate training of the machine learning model, generating a synchronized training dataset mirroring the training dataset stored in the virtual object storage service; storing the synchronized training dataset in a second local storage device; and fetching training data from the synchronized training dataset stored in the second local storage device as the training of the machine learning model is executed.

Abstract: Systems for cooling one or more electronic devices are disclosed, each of the one or more electronic devices including one or more heat-generating components. An immersion cooling includes a first container adapted for receiving a first heat-transfer liquid, the one or more electronic devices being, at least in part, immersed in the first heat-transfer liquid such that the first heat-transfer liquid collects, in use, at least a portion of a thermal energy generated by the one or more heat-generating components and a second container adapted for receiving a second heat-transfer liquid, the first container being, at least in part, immersed in the second container cooling system.

Abstract: The disclosed systems and methods are directed to providing the mobile robot-assisted thermal monitoring of server racks in a datacenter. The thermal monitoring comprises a plurality of thermal sensors to detect temperature data of each of the server racks, a plurality of power distribution units (PDUs) to detect electrical power consumption of each of the servers, a datacenter operations controller configured to generate a temperature gradient profile over time for each of the server racks, determine a potential anomalous thermal condition of a server rack, and identify the server rack as a point-of-interest (POI). The thermal monitoring further comprises a mobile robot configured to travel to and capture the surface temperature image data of the identified POI server rack based on a provided navigation transit route and the datacenter operations controller updating the temperature gradient profile based on the capture the surface temperature image data.

Abstract: A datacenter dry cooling system for cooling a heat generating source is presented that incorporates a first closed loop arrangement including an air-to-liquid heat exchanger to re-cool heat imparted onto the first cooling fluid, and a first pump that forwards the re-cooled first cooling fluid to the heat generating source. A second cooling fluid open loop arrangement that conveys a second cooling fluid via a second pump to supply the second cooling fluid onto an evaporating pad of an air-to-liquid heat exchanger for cooling dampening. The system also incorporates at least one fan to cause an air flow through the evaporating pad and the first air-to-liquid heat exchanger along with an electronic controller to manage and control the flow rates of the first and second pumps, the flow rate of the second cooling fluid valve, and/or a rotation speed of the at least one fan.

Abstract: The present invention describes an intelligent power management system designed for optimizing energy use in data center server assemblies. It can use various coolants, like liquid or gas, and adjusts to changing processor loads. The system integrates a coolant loop, a variable flow coolant circulator, and an advanced control unit. The advanced control unit dynamically regulates coolant flow, server temperature, and total power consumption, and receives temperature data from each server's baseboard management controller and power data from a power distribution unit. The advanced control unit computes the best coolant flow rate to minimize total power use while keeping server temperature below a safety limit, and uses either a static or a machine learning model with reinforcement learning and a reward mechanism to enhance real-time calculations and overall efficiency over time.

Abstract: System and method for providing electric power to an electronic device. The method includes receiving electric power at a corresponding voltage, receiving a first standby electric power at a first standby voltage, directing the electric power to the electronic device through a corresponding fuse state indicator selectively connecting a corresponding power input of a monitoring circuit to the electronic device and relying on the first standby electric power to operate, combining the fuse state signal with a signal indicative of a presence of the electric power at the power input of the monitoring circuit to form a main state signal that can be selectively an alive state or a failure state; and, in response to the main state signal being in a failure state, disconnecting the electronic device from the power source.

Abstract: Methods and systems for detecting Denial-of-Service (DoS) attacks on a network. The method includes clustering connection requests received at the network during a first collection time having a first time span based on communication datasets extracted therefrom. For at least one connection cluster, a first DoS attack detection algorithm is executed to determine a first probability that the connection cluster is associated with a DoS attack. The first confidence score is compared with a first threshold to determine whether a mitigation order is to be generated. The connection cluster is adjusted by extracting second communication datasets from connection requests received at the network during a second collection time having a second time span longer than the first time span, and a second DoS attack detection algorithm is executed to determine a second probability that the connection cluster is associated with a DoS attack.

Abstract: There is disclosed a method and system for identifying the effects of an item failing in a network. Information about a first item in a network is received. The information indicates that the first item provides a service to other items in the network. A graph of the network is retrieved. A node in the graph that corresponds to the first item is identified. Edges are added to the graph to connect the node of the first item to nodes of the other items that receive the service from the first item. A request is received to model a failure of the first item. The nodes connected to the node that corresponds to the first item are retrieved. Operational rules are applied to each of the retrieved nodes to identify if any of their corresponding items would fail. A list of items that would fail is output.

Abstract: Method and system for detecting a Denial-of-Service (DoS) attack on a network. The method includes sampling a traffic carried by the network between data processing devices, the traffic comprising data packets transmitted between the data processing devices, determining statistically estimated features of the traffic based on the sampling thereof, executing one or more heuristic algorithm based on the statistically estimated features, the one or more heuristic algorithms being configured to generate a confidence score indicative of a probability that at least some of the data packets constitute a DoS attack and comparing the confidence score with a confidence threshold to determine whether a mitigation order is to be generated.

Abstract: A system for configuring with an overlay configuration, a datacenter network of networking devices comprising at least an underlay fabric, an edge, and a ToR switch interconnecting one or several hosts on a rack, communicably connected to each other with an underlay configuration. The system comprises an API providing a user with a set of primitives providing instantiation capability of respective generic networking resources, the primitives being configured to be invoked and assembled by the user so as to provide an abstracted view of the overlay configuration. The system further comprises a Workflow engine to compute the overlay configuration according to the abstracted view, and an Overlay configuration builder to push the overlay configuration to the ToR switch and the edge. The pushed overlay configuration configures a NVE in the datacenter network as one of the ToR switch, depending on the business use case of the datacenter associated with the identity of the user.

Abstract: A method for sharing a source dataset between a first user device and a second user device, the method being executed by a first processor communicably connected to the first user device. The method includes accessing a configuration file having instruction sets enabling retrieval of source datasets stored respectively in corresponding data storage services; receiving an indication from the first user device including a selected source dataset from the source datasets, and a corresponding selected data storage service from the data storage services; extracting an instruction set corresponding to the selected source dataset; executing the extracted instruction set, and retrieving the selected source dataset onto the first user device; and transferring the configuration file to the second user device to enable the second user device to, when executing the extracted instruction set, retrieve the selected source dataset from the selected data storage service.

Abstract: A method and system allow to store and retrieve data values and data structure in a memory of an embedded system, the memory also storing executable code uploaded to the embedded system by a first external system through a first version of a firmware. A serializer is configured to receive from the executable code, sizes of data, corresponding to a number of chunks of the data, together with values of the chunks, to form a series of bytes therewith through concatenation, to calculate a total size of the series of bytes, and to write to the memory, successively the total size and the formed series of bytes. A deserializer is configured to read from the memory, successively a total size and a series of bytes, to deconcatenate the series of bytes into successively sizes of data, corresponding to a number of chunks of the data, together with values of the chunks, and to send the same to the executable code.

Abstract: A cooling block includes a body having a thermal transfer surface configured to be in thermal contact with a heat-generating electronic component. The body defines a fluid conduit having a plurality of channels such that the cooling fluid flows in parallel along the channels. Each channel is defined between first and second internal sidewalls; and top and bottom walls. Each channel has either: a first dimensional configuration including first height and width values; or a second dimensional configuration including second height and width values. The first height value is greater than the second height value. The second width value is greater than the first width value. A thickness of a material of the body between the thermal transfer surface and a first channel is less than a thickness of the material of the body between the thermal transfer surface and a second channel.

Abstract: A cooling system and a method for cooling an electronic device are disclosed. The electronic device includes a heat-generating component. The cooling arrangement includes an immersion casing that contains a dielectric cooling liquid, the electronic device being, at least in part, immersed in the dielectric cooling liquid such that the dielectric cooling liquid collects, in use, at least a first portion of a thermal energy generated by the heat-generating component, and a photovoltaic device proximate to the immersion casing and configured to convert at least in part a second portion of the thermal energy generated by the heat-generating component into electric energy, the second portion having radiated from the dielectric cooling liquid and through the immersion casing.

Abstract: A cooling assembly and method for cooling a plurality of heat-generating components. The cooling assembly includes a main liquid inlet for receiving a heat-transfer liquid, a main liquid outlet for discharging the heat-transfer liquid and a plurality of liquid cooling units. Each liquid cooling unit is in thermal contact with a corresponding heat-generating component and includes an internal liquid conduit, the heat-transfer liquid flowing in the internal liquid conduit collecting thermal energy from the corresponding heat-generating component. The liquid cooling units are arranged in a plurality of groups, each group comprising one or more liquid cooling units fluidly connected in parallel to one another. The groups are arranged in at least one cluster, each cluster comprising two or more groups fluidly connected in series.

Abstract: System and method for storing a captured aerial image on a data storage device, the captured aerial image being associated with a corresponding metadata file. The method including accessing the captured aerial image and the corresponding metadata file, extracting a sequence of stored aerial images from a compressed file stored on the data storage device, the sequence of stored aerial images comprising at least one stored aerial image, determining a first position for inserting the captured aerial image in the sequence of stored aerial images, updating the sequence of stored aerial images by inserting the captured aerial image in the sequence of stored aerial images at the first position, updating the compressed file by applying an encoding algorithm to the sequence of stored aerial images and storing the compressed file on the data storage device.

Abstract: A method for controlling a cooling system of a rack, the rack comprising a heat generating component. The method comprises receiving, by a controller, at least one first temperature indications indicative of temperature variations of at least one heat transfer fluid circulating in at least one respective liquid channel of the cooling system; receiving, by the controller, at least one second temperature indications indicative of a temperature of an air flow of ambient air within the rack, the air flow being generated by at least one fan of the cooling system; and adjusting, based on the at least one first and second temperature indications, a rotational speed of the at least one fan and a rotational speed of at least one pump configured for causing the at least one heat transfer fluid to flow in at least one respective liquid channel.

Abstract: A cooling block assembly for cooling a heat-generating electronic component includes: an upper block portion defining at least one internal fluid conduit and having a lower surface configured to face toward the heat-generating electronic component; a boss spaced from the lower surface and having a thermal transfer surface configured to be in thermal contact with the heat-generating electronic component, the thermal transfer surface being offset from the lower surface, a periphery of the thermal transfer surface being smaller than a periphery of the upper block portion, the periphery of the thermal transfer surface being contained within the periphery of the upper block portion in a projection thereof on a plane parallel to the thermal transfer surface; and heat distributing devices for distributing heat through a phase change of a working substance contained therein, each heat distributing device being disposed partially between the boss and the lower surface.

Abstract: An immersive cooling system is described. The system includes an immersive cooling container, including: a first reservoir; a second reservoir; numerous slots, each configured to hold a casing; a reservoir connector corresponding to each slot, and configured to provide fluid communication with the first reservoir; and a pump configured to convey a dielectric immersion cooling liquid from the second reservoir to the first reservoir. The immersive cooling system also includes a casing, configured to contain an electronic device and to fit within a slot of the container. The casing includes an inlet configured to be in fluid communication with the reservoir connector of the slot within which the casing is disposed to facilitate flow of the cooling liquid into an interior of the casing through the reservoir connector and an outlet configured to facilitate flow of the cooling liquid from the interior of the casing into the second reservoir.

Abstract: An immersion tank storage system for a data center, includes: a frame defining a plurality of storage levels disposed above one another; a plurality of immersion tanks supported by the frame, each immersion tank configured to contain electronic devices and an immersion cooling liquid for cooling thereof, each storage level being configured to house first and second immersion tanks; and means for synchronously moving the first and second immersion tanks housed in at least one of the storage levels between: a storage position in which the first and second immersion tanks are spaced from each other by a first distance; and an access position in which the first and second immersion tanks are spaced from each other by a second distance greater than the first distance, wherein in the access position, the first and second immersion tanks are accessible by an operator for accessing the electronic devices contained therein.