Advanced Micro Devices Corporation

Ati Technologies Ulc

Ann M Ling

Vydhyanathan Kalyanasundharam

Ganesh Balakrishnan

Amit P Apte

Eric Christopher Morton

Benjamin Tsien

Kevin Michael Lepak

John P Petry

Richard Martin Born

Bryan P Broussard

Michael J Tresidder

Ivan Yanfeng Wang

Alexander J Branover

Maurice B Steinman

Ravindra Nath Bhargava

Ashish Jain

A coherent memory fabric includes a plurality of coherent master controllers and a coherent slave controller. The plurality of coherent master controllers each include a response data buffer. The coherent slave controller is coupled to the plurality of coherent master controllers. The coherent slave controller, responsive to determining a selected coherent block read command is guaranteed to have only one data response, sends a target request globally ordered message to the selected coherent master controller and transmits responsive data. The selected coherent master controller, responsive to receiving the target request globally ordered message, blocks any coherent probes to an address associated with the selected coherent block read command until receipt of the responsive data is acknowledged by a requesting client.

Coherent block read fulfillment

A method includes, in a cache directory, storing an entry associating a memory region with an exclusive coherency state, and in response to a memory access directed to the memory region, transmitting a demote superprobe to convert at least one cache line of the memory region from an exclusive coherency state to a shared coherency state.

Page rinsing scheme to keep a directory page in an exclusive state in a single complex

Disclosed are examples of a system and method to communicate cache line eviction data from a CPU subsystem to a home node over a prioritized channel and to release the cache subsystem early to process other transactions.

Multi-level cache coherency protocol for cache line evictions

Systems, apparatuses, and methods for reducing chiplet interrupt latency are disclosed. A system includes one or more processing nodes, one or more memory devices, a communication fabric coupled to the processing unit(s) and memory device(s) via link interfaces, and a power management unit. The power management unit manages the power states of the various components and the link interfaces of the system. If the power management unit detects a request to wake up a given component, and the link interface to the given component is powered down, then the power management unit sends an out-of-band signal to wake up the given component in parallel with powering up the link interface. Also, when multiple link interfaces need to be powered up, the power management unit powers up the multiple link interfaces in an order which complies with voltage regulator load-step requirements while minimizing the latency of pending operations.

Reducing chiplet wakeup latency

Systems, apparatuses, and methods for accelerating cache to cache data transfers are disclosed. A system includes at least a plurality of processing nodes and prediction units, an interconnect fabric, and a memory. A first prediction unit is configured to receive memory requests generated by a first processing node as the requests traverse the interconnect fabric on the path to memory. When the first prediction unit receives a memory request, the first prediction unit generates a prediction of whether data targeted by the request is cached by another processing node. The first prediction unit is configured to cause a speculative probe to be sent to a second processing node responsive to predicting that the data targeted by the memory request is cached by the second processing node. The speculative probe accelerates the retrieval of the data from the second processing node if the prediction is correct.

Cache to cache data transfer acceleration techniques

An apparatus and method for dynamically adjusting power limits for processing nodes and other components, such as peripheral interfaces, is disclosed. The apparatus includes multiple processing nodes and other components, and further includes a power management unit configured to set a first frequency limit for at least one of the processing nodes responsive to receiving an indication of a first detected temperature greater than a first temperature threshold. Initial power limits are set below guard-band power limits for components that do not have reliable reporting of power consumption or for cost or power saving reasons. The amount of throttling of processing nodes is used to adjust the power limits for the processing nodes and these components.

Method for adaptive performance optimization of the soc

Growing community of inventors

Folsom, CA, United States of America

Ann M Ling