Abstract: The present disclosure provides for a converged compute platform architecture, including a first infrastructure processing unit (IPU)-only configuration and a second configuration wherein the IPU is coupled to a central processing unit, such as an x86 processor. Connectivity between the two configurations may be accomplished with a PCIe switch, or the two configurations may communicate through remote direct memory access (RDMA) techniques. Both configurations may use ML acceleration through a single converged architecture.

Abstract: Compressing connection state information for a network connection including receiving an input bitmap having a sequence of bits describing transmit states and receive states; partitioning the input bitmap into a plurality of equal size blocks; partitioning each of the blocks into a plurality of equal sized sectors; generating a block valid sequence indicating the blocks having at least one bit set; generating, for each block having at least one bit set, a sector information sequence, the sector information sequence indicating, for the corresponding block, the sectors that have at least one bit set and an encoding type for each sector; and generating one or more symbols by encoding each sector that has at least one bit set.

Abstract: The present disclosure provides for a converged compute platform architecture, including a first infrastructure processing unit (IPU)-only configuration and a second configuration wherein the IPU is coupled to a central processing unit, such as an x86 processor. Connectivity between the two configurations may be accomplished with a PCIe switch, or the two configurations may communicate through remote direct memory access (RDMA) techniques. Both configurations may use ML acceleration through a single converged architecture.

Abstract: Methods, systems, and apparatus, including computer-readable storage media for resource isolation between connections with shared hardware resources. A network device, such as a network interface card, is configured to determine dynamic resource limits for each connection, and backpressure each connection individually to avoid a global pause when the shared hardware resources are oversubscribed by the current connections. As a result, slower connections may be paused for exceeding resource limits, protecting faster connections from slowing down because resources are shared between both types of connections. Dynamic resource limits can be generated and updated not only per connection, but also based on subsets of the shared hardware resources assigned to different sources of data, as well assigned to different types of transactions communicated over a connection. A hardware-assisted transport layer can be configured to apply dynamic resource limits individually to different connections.

Abstract: Aspects of the disclosure are directed to establishing and utilizing multiple flows, e.g., data paths, within a single connection between two end points in a network. Packets being transmitted between the endpoints can be load-balanced among multiple flows using a set of flow labels. The flow label is determined using scheduling logic. The flow labels include a flow weight that encodes how the packet is mapped to a given flow. The flow weight may be used to determine a congestion window for each flow in the connection. As packets are communicated between the endpoints, congestion control data and acknowledgement coalescing entries are updated before an acknowledgement is sent. Each flow maintains a counter of the number of acknowledgments received. The number of acknowledgments received is used to implement congestion control.

Abstract: Systems and methods of offloading multicast virtual network packet processing to a network interface card are provided. In an example implementation, a network interface card can route packets in a virtual network. The network interface card can be configured to receive a data packet having a multicast header for transmission to a plurality of destination virtual machines. The network interface card can retrieve a list of next hop destinations for the data packet. The network interface card can replicate the packet for each next hop destination. The network interface card can encapsulate each replicated packet with a unicast header that includes a next hop destination virtual IP address indicating the next hop destination and a source virtual IP address, and transmit the encapsulated packets.

Abstract: A custom processor core is provided, wherein the custom processor core may be used for congestion control in reliable transport protocols. The hardware architecture of the custom processor core allows for custom instructions, special register sets, and datapath enhancements for accelerating congestion control algorithms to achieve higher performance.

Abstract: An application specific integrated circuit (ASIC) is provided for reliable transport of packets. The network interface card may include a reliable transport accelerator (RTA). The RTA may include a cache lookup database. The RTA may be configured to determine, from a received data packet, a connection identifier and query the cache lookup database for a cache entry corresponding to a connection context having the connection identifier. In response to the query, the RTA may receive a cache hit or a cache miss.

Abstract: A communication technology that provides for handling of failed packet transmissions to reduce retransmission attempts and uses resynchronization to prevent tearing down of connections. Thereby, providing for more resilient connections. In an implementation, an initiator entity may determine that a negative acknowledgment indicates that an operation for a particular packet is completed in error by a target entity, and transmit to the target entity a resynchronization packet without tearing down the connection.

Abstract: A communication protocol system is provided for reliable transport of packets. A content addressable memory hardware architecture including an acknowledgment coalescing module in communication with a content addressable memory (CAM). The acknowledgment coalescing module coalesces multiple acknowledgement packets as a single acknowledgement packet to reduce the overall numbers of the packet transmission in the communication protocol system. In addition, the acknowledgment coalescing module may also provide a piggyback mechanism to carry acknowledge information in a regular data packet. Thus, the need to generate a new acknowledgement packet may be eliminated. Accordingly, the network congestion and latency may be reduced, and the communication and transmission efficiency are enhanced.

Abstract: A system includes a first processor configured to analyze packets received over a communication protocol system and determine one or more congestion indicators from the analysis of the data packets, the one or more congestion indicators being indicative of network congestion for data packets transmitted over a reliable transport protocol layer of the communication protocol system. The system also includes a rate update engine separate from the packet datapath and configured to operate a second processor to receive the determined one or more congestion indicators, determine one or more congestion control parameters for controlling transmission of data packets based on the received one or more congestion indicators, and output a congestion control result based on the determined one or more congestion control parameters.

Abstract: A flow table management system can include a hardware memory module communicatively coupled to a network interface card. The hardware memory module is configured to store a flow table including a plurality of network flow entries. The network interface card further includes a flow table age cache configured to store a set of recently active network flows and a flow table management module configured to manage a duration for which respective network flow entries in the flow table stored in the hardware memory module remain in the flow table using the flow table age cache. In some implementations, age information about each respective flow in the flow table is stored in the hardware memory module in an age state table that is separate from the flow table.

Abstract: Compressing connection state information for a network connection including receiving an input bitmap having a sequence of bits describing transmit states and receive states; partitioning the input bitmap into a plurality of equal size blocks; partitioning each of the blocks into a plurality of equal sized sectors; generating a block valid sequence indicating the blocks having at least one bit set; generating, for each block having at least one bit set, a sector information sequence, the sector information sequence indicating, for the corresponding block, the sectors that have at least one bit set and an encoding type for each sector; and generating one or more symbols by encoding each sector that has at least one bit set.

Abstract: A communication protocol system is provided for reliable transport of packets. In this regard, an initiator entity may determine that outgoing data is to be transmitted to a target entity. The initiator entity may transmit, to the target entity, a solicited push request requesting the outgoing data to be placed at the target entity. In response to the solicited push request, the initiator entity may receive a push grant from the target entity. In response to the push grant, the initiator entity may transmit to the target entity the outgoing data to be placed at the target entity.

Abstract: The present disclosure provides for a converged compute platform architecture, including a first infrastructure processing unit (IPU)-only configuration and a second configuration wherein the IPU is coupled to a central processing unit, such as an x86 processor. Connectivity between the two configurations may be accomplished with a PCIe switch, or the two configurations may communicate through remote direct memory access (RDMA) techniques. Both configurations may use ML acceleration through a single converged architecture.

Abstract: A network interface card with traffic shaping capabilities and methods of network traffic shaping with a network interface card are provided. The network interface card and method can shape traffic originating from one or more applications executing on a host network device. The applications can execute in a virtual machine or containerized computing environment. The network interface card and method can perform or include several traffic shaping mechanisms including, for example and without limitation, a delayed completion mechanism, a time-indexed data structure, a packet builder, and a memory manager.

Abstract: Systems and methods of offloading multicast virtual network packet processing to a network interface card are provided. In an example implementation, a network interface card can route packets in a virtual network. The network interface card can be configured to receive a data packet having a multicast header for transmission to a plurality of destination virtual machines. The network interface card can retrieve a list of next hop destinations for the data packet. The network interface card can replicate the packet for each next hop destination. The network interface card can encapsulate each replicated packet with a unicast header that includes a next hop destination virtual IP address indicating the next hop destination and a source virtual IP address, and transmit the encapsulated packets.

Abstract: A system includes a first processor configured to analyze packets received over a communication protocol system and determine one or more congestion indicators from the analysis of the data packets, the one or more congestion indicators being indicative of network congestion for data packets transmitted over a reliable transport protocol layer of the communication protocol system. The system also includes a rate update engine separate from the packet datapath and configured to operate a second processor to receive the determined one or more congestion indicators, determine one or more congestion control parameters for controlling transmission of data packets based on the received one or more congestion indicators, and output a congestion control result based on the determined one or more congestion control parameters.

Abstract: A system is provided for synchronizing clocks. The system includes a plurality of devices in a network, each device having a local clock. The system is configured to synchronize the local clocks according to a primary spanning tree, where the primary spanning tree has a plurality of nodes connected through a plurality of primary links, each node of the plurality of nodes representing a respective device of the plurality of devices. The system is also configured to compute a backup spanning tree before a failure is detected in the primary spanning tree, wherein the backup spanning tree includes one or more backup links that are different from the primary links. As such, upon detection of a failure in the primary spanning tree, the system reconfigures the plurality of devices such that clock synchronization is performed according to the backup spanning tree.

Abstract: A system includes a first processor configured to analyze packets received over a communication protocol system and determine one or more congestion indicators from the analysis of the data packets, the one or more congestion indicators being indicative of network congestion for data packets transmitted over a reliable transport protocol layer of the communication protocol system. The system also includes a rate update engine separate from the packet datapath and configured to operate a second processor to receive the determined one or more congestion indicators, determine one or more congestion control parameters for controlling transmission of data packets based on the received one or more congestion indicators, and output a congestion control result based on the determined one or more congestion control parameters.