Abstract: Systems, apparatuses, and methods for implementing a cancel and replay mechanism for ordered requests are disclosed. A system includes at least an ordering master, a memory controller, a coherent slave coupled to the memory controller, and an interconnect fabric coupled to the ordering master and the coherent slave. The ordering master generates a write request which is forwarded to the coherent slave on the path to memory. The coherent slave sends invalidating probes to all processing nodes and then sends an indication that the write request is globally visible to the ordering master when all cached copies of the data targeted by the write request have been invalidated. In response to receiving the globally visible indication, the ordering master starts a timer. If the timer expires before all older requests have become globally visible, then the write request is cancelled and replayed to ensure forward progress in the fabric and avoid a potential deadlock scenario.

Abstract: A system for implementing load balancing schemes includes one or more processing units, a memory, and a communication fabric with a plurality of switches coupled to the processing unit(s) and the memory. A switch of the fabric determines a first number of streams on a first input port that are targeting a first output port. The switch also determines a second number of requestors, from all input ports, that are targeting the first output port. Then, the switch calculates a throttle factor for the first input port by dividing the first number of streams by the second number of streams. The switch applies the throttle factor to regulate bandwidth on the first input port for requestors targeting the first output port. The switch also calculates throttle factors for the other ports and applies the throttle factors when regulating bandwidth on the other ports.

Abstract: A method for steering data for an I/O write operation includes, in response to receiving the I/O write operation, identifying, at an interconnect fabric, a cache of one of a plurality of compute complexes as a target cache for steering the data based on at least one of: a software-provided steering indicator, a steering configuration implemented at boot initialization, and coherency information for a cacheline associated with the data. The method further includes directing, via the interconnect fabric, the identified target cache to cache the data from the I/O write operation. The data is temporarily buffered at the interconnect fabric, and if the target cache attempts to fetch the data while the data is still buffered at the interconnect fabric, the interconnect fabric provides a copy of the buffered data in response to the fetch operation instead of initiating a memory access operation to obtain the data from memory.

Abstract: Systems, apparatuses, and methods for implementing a cancel and replay mechanism for ordered requests are disclosed. A system includes at least an ordering master, a memory controller, a coherent slave coupled to the memory controller, and an interconnect fabric coupled to the ordering master and the coherent slave. The ordering master generates a write request which is forwarded to the coherent slave on the path to memory. The coherent slave sends invalidating probes to all processing nodes and then sends an indication that the write request is globally visible to the ordering master when all cached copies of the data targeted by the write request have been invalidated. In response to receiving the globally visible indication, the ordering master starts a timer. If the timer expires before all older requests have become globally visible, then the write request is cancelled and replayed to ensure forward progress in the fabric and avoid a potential deadlock scenario.

Abstract: A system for automatically discovering fabric topology includes at least one or more processing units, one or more memory devices, a security processor, and a communication fabric with an unknown topology coupled to the processing unit(s), memory device(s), and security processor. The security processor queries each component of the fabric to retrieve various attributes associated with the component. The security processor utilizes the retrieved attributes to create a network graph of the topology of the components within the fabric. The security processor generates routing tables from the network graph and programs the routing tables into the fabric components. Then, the fabric components utilize the routing tables to determine how to route incoming packets.

Abstract: Systems, apparatuses, and methods for maintaining a region-based cache directory are disclosed. A system includes multiple processing nodes, with each processing node including a cache subsystem. The system also includes a cache directory to help manage cache coherency among the different cache subsystems of the system. In order to reduce the number of entries in the cache directory, the cache directory tracks coherency on a region basis rather than on a cache line basis, wherein a region includes multiple cache lines. Accordingly, the system includes a region-based cache directory to track regions which have at least one cache line cached in any cache subsystem in the system. The cache directory includes a reference count in each entry to track the aggregate number of cache lines that are cached per region. If a reference count of a given entry goes to zero, the cache directory reclaims the given entry.

Abstract: A method for steering data for an I/O write operation includes, in response to receiving the I/O write operation, identifying, at an interconnect fabric, a cache of one of a plurality of compute complexes as a target cache for steering the data based on at least one of: a software-provided steering indicator, a steering configuration implemented at boot initialization, and coherency information for a cacheline associated with the data. The method further includes directing, via the interconnect fabric, the identified target cache to cache the data from the I/O write operation. The data is temporarily buffered at the interconnect fabric, and if the target cache attempts to fetch the data while the data is still buffered at the interconnect fabric, the interconnect fabric provides a copy of the buffered data in response to the fetch operation instead of initiating a memory access operation to obtain the data from memory.

Abstract: A system for implementing load balancing schemes includes one or more processing units, a memory, and a communication fabric with a plurality of switches coupled to the processing unit(s) and the memory. A switch of the fabric determines a first number of streams on a first input port that are targeting a first output port. The switch also determines a second number of requestors, from all input ports, that are targeting the first output port. Then, the switch calculates a throttle factor for the first input port by dividing the first number of streams by the second number of streams. The switch applies the throttle factor to regulate bandwidth on the first input port for requestors targeting the first output port. The switch also calculates throttle factors for the other ports and applies the throttle factors when regulating bandwidth on the other ports.

Abstract: Systems, apparatuses, and methods for dynamic buffer management in multi-client token flow control routers are disclosed. A system includes at least one or more processing units, a memory, and a communication fabric with a plurality of routers coupled to the processing unit(s) and the memory. A router servicing multiple active clients allocates a first number of tokens to each active client. The first number of tokens is less than a second number of tokens needed to saturate the bandwidth of each client to the router. The router also allocates a third number of tokens to a free pool, with tokens from the free pool being dynamically allocated to different clients. The third number of tokens is equal to the difference between the second number of tokens and the first number of tokens. An advantage of this approach is reducing the amount of buffer space needed at the router.

Abstract: Systems, apparatuses, and methods for implementing priority adjustment forwarding are disclosed. A system includes at least one or more processing units, a memory, and a communication fabric coupled to the processing unit(s) and the memory. The communication fabric includes a plurality of arbitration points. When a client determines that its bandwidth requirements are not being met, the client generates and sends an in-band priority adjustment request to the nearest arbitration point. This arbitration point receives the in-band priority adjustment request and then identifies any pending requests which are buffered at the arbitration point which meet the criteria specified by the in-band priority adjustment request. The arbitration point adjusts the priority of any identified requests, and then the arbitration point forwards the in-band priority adjustment request on the fabric to the next upstream arbitration point which processes the in-band priority adjustment request in the same manner.

Abstract: A system and method for network traffic management between multiple nodes are described. A computing system includes multiple nodes connected to one another. When a home node determines a number of nodes requesting read access for a given data block assigned to the home node exceeds a threshold and a copy of the given data block is already stored at a first node of the multiple nodes in the system, the home node sends a command to the first node. The command directs the first node to forward a copy of the given data block to the home node. The home node then maintains a copy of the given data block and forwards copies of the given data block to other requesting nodes until the home node detects a write request or a lock release request for the given data block.

Abstract: A technique for source-side memory request network admission control includes adjusting, by a first node, a rate of injection of memory requests by the first node into a network coupled to a memory system. The adjusting is based on an injection policy for the first node and memory request efficiency indicators. The method may include injecting memory requests by the first node into the network coupled to the memory system. The injecting has the rate of injection. The technique includes adjusting the rate of injection by the first node. The first node adjusts the rate of injection according to an injection policy for the first node and memory request efficiency indicators. The injection policy may be based on an injection rate limit for the first node. The injection policy for the first node may be based on an injection rate limit per memory channel for the first node.

Abstract: A system and method for network traffic management between multiple nodes are described. A computing system includes multiple nodes connected to one another. When a home node determines a number of nodes requesting read access for a given data block assigned to the home node exceeds a threshold and a copy of the given data block is already stored at a first node of the multiple nodes in the system, the home node sends a command to the first node. The command directs the first node to forward a copy of the given data block to the home node. The home node then maintains a copy of the given data block and forwards copies of the given data block to other requesting nodes until the home node detects a write request or a lock release request for the given data block.

Abstract: A technique for source-side memory request network admission control includes adjusting, by a first node, a rate of injection of memory requests by the first node into a network coupled to a memory system. The adjusting is based on an injection policy for the first node and memory request efficiency indicators. The method may include injecting memory requests by the first node into the network coupled to the memory system. The injecting has the rate of injection. The technique includes adjusting the rate of injection by the first node. The first node adjusts the rate of injection according to an injection policy for the first node and memory request efficiency indicators. The injection policy may be based on an injection rate limit for the first node. The injection policy for the first node may be based on an injection rate limit per memory channel for the first node.