Mellanox Technologies, Ltd.

Yair Chasdai

Liran Daniel

David Daniel Pilnik

Gary Mataev

Maty Golovaty

Ran Sandhaus

Alon Riesenberg

Noam Rahoum

Ido Zemah

A peripheral device includes a bus interface, a first processor and a second processor. The bus interface is to communicate over a peripheral bus. The first processor is to manage communication over the peripheral bus by executing bus-maintenance software code, the bus-maintenance software code being executed from one or more first layers of a multi-layer memory. The second processor is to update the bus-maintenance software code from an existing version to an updated version, by (i) loading the updated version to one or more second layers of the multi-layer memory, higher in hierarchy than the one or more first layers, and (ii) invalidating the existing version in the one or more first layers, thereby forcing fetching of the updated version from the one or more second layers to the one or more first layers and to start executing the updated version.

Fast in-service software updating using multi-layer memory

A system for cycle-based per-flow packet drop counting. The system comprises a communication network which interconnects nodes including receiver/s, supervisor/s, and sender/s. The sender sends flow/s of packets, via link/s, to the receiver/s. Flow/s are divided into cycles whose lengths are known to sender, receiver and supervisor. The sender adds to each packet indications of: a cycle during which the packet was sent, and whether the packet belongs to a given flow for which packet drops are being counted. The sender reports to the supervisor a number of packets sent during the individual cycle. The receiver reports to the supervisor, by counting indications, for the individual flow, how many packets, sent during the individual cycle, were received. The supervisor computes a difference between the number of packets sent and received, and, for certain difference value/s, implements further action/s.

Cycle-based per-flow packet drop counting

Devices and methods to identify malicious usage of a network device. In at least one embodiment, a network device comprises circuitry for performing a networking function and collecting telemetry data indicative of the performance of the networking function. The network device obtains an inference of a network traffic pattern using a machine learning model, and responds to the inference.

Network devices assisted by machine learning

In one embodiment, a device includes a data compression engine configured to process data packets of multiple network flows received from a network, and compress respective groups of the received data packets into respective compressed packets for sending over the network, each respective compressed packet including respective ones of the received data packets from different ones of the network flows, wherein the network interface is configured to send the compressed packets over the network.

Multi-flow compression

A network device includes at least one port, a memory, data-path circuitry, and a processor. The at least one port is to exchange packets with a network. The data-path circuitry is to process the packets. The memory is to store signatures of traffic patterns according to a locality-sensitive signature function., and corresponding parameter settings for the data-path circuitry. The processor is to assess a current traffic pattern of the packets, to calculate a current signature over the current traffic pattern using the locality-sensitive signature function, to query the memory using the current signature, to configure the data-path circuitry, in response to finding a stored signature that is within a specified distance from the current signature, with a parameter setting that corresponds to the found signature, and take an alternative action in response to finding that no stored signature is within the specified distance from the current signature.

Growing community of inventors

Tel Aviv, Israel

Yair Chasdai