Abstract: Described herein is a network fabric including a plurality of graphical processing unit (GPU) clusters. The plurality of GPU clusters includes at least a first GPU cluster operating at a first speed and a second GPU cluster operating at a second speed that is different than the first speed. The network fabric includes a plurality of blocks, wherein each block includes: (a) one or more racks that host a GPU cluster, and (b) a plurality of switches arranged in a hierarchical structure that communicatively couple the block to other blocks included in the network fabric. Responsive to receiving a request to execute a workload, allocating one or more GPUs from the plurality of GPU clusters to execute the workload.

Abstract: A plurality of GPU clusters are communicatively coupled with one another via a plurality of network devices arranged in a hierarchical structure, wherein the GPU clusters includes at least a first GPU cluster operating at a first speed and a second GPU cluster operating at a second speed that is different than the first speed. A routing policy is configured for each network device, wherein the configuring includes establishing a mapping of each incoming port-link of the network device to a unique outgoing port-link of the network device. For a packet transmitted by a GPU of a host machine and received by a first network device, an incoming port-link of the first network device is determined on which the packet was received and based on the configuring, an outgoing port-link is identified that corresponds to the incoming port-link. The packet is forwarded on the outgoing port-link of the network device.

Abstract: Discussed herein is a framework that provisions for customized processing for different classes of traffic. A network device in a communication path between a source host machine and a destination host machine extracts a tag from a packet received by the network device. The packet originates at a source executing on the source host machine and whose destination is the destination host machine. The tag set by the source and indicative of a first traffic class to be associated with the packet, the first traffic class being selected by the source from a plurality of traffic classes. The network device determines, based on the tag, that the first traffic class corresponds to a latency sensitive traffic and processes the packet using one or more settings configured at the network device for processing packets associated with the first traffic class.

Abstract: Discussed herein is a framework that provisions for customized processing for different classes of traffic. A network device in a communication path between a source host machine and a destination host machine extracts a tag from a packet received by the network device. The packet originates at a source executing on the source host machine and whose destination is the destination host machine. The tag set by the source and indicative of a first traffic class to be associated with the packet, the first traffic class being selected by the source from a plurality of traffic classes. The network device determines, based on the tag, that the first traffic class corresponds to a bandwidth sensitive traffic and processes the packet using one or more settings configured at the network device for processing packets associated with the first traffic class.

Abstract: Discussed herein is a framework that provisions for customized processing for different classes of traffic. A network device in a communication path between a source host machine and 5 a destination host machine extracts a tag from a packet received by the network device. The packet originates at a source executing on the source host machine and whose destination is the destination host machine. The tag set by the source and indicative of a first traffic class to be associated with the packet, the first traffic class being selected by the source from a plurality of traffic classes. The network device determines the first traffic class based on the tag extracted from the packet and 10 processes the packet based on the first traffic class.

Abstract: Discussed herein is a routing mechanism for graphical processing units (GPUs) that are hosted on several host machines in a cloud environment. For a packet transmitted by a GPU of a host machine and received by a network device, the network device determines an incoming port-link of the network device on which the packet was received. The network device obtains a flow information associated with the packet, and computes, based on the flow information, an outgoing port-link of the network device in accordance with a hashing algorithm. The hashing algorithm is configured to hash packets received on a particular incoming port-link of the network device to be transmitted on a same outgoing port-link of the network device. The network device forwards the packet on the outgoing port-link of the network device.

Abstract: Discussed herein is a routing mechanism for graphical processing units (GPUs) that are hosted on several host machines in a cloud environment. For a packet transmitted by a GPU of a host machine and received by a network device, the network device determines an incoming port-link of the network device on which the packet was received. The network devices identifies, based on a GPU routing policy, an outgoing port-link of the network device that corresponds to the incoming port-link. The GPU routing policy is preconfigured prior to receiving the packet and establishes a mapping of each incoming port-link of the network device to a unique outgoing port-link of the network device. The packet is forwarded on the outgoing port-link of the network device.