SCHEDULING TRAINING OF AN INTER-CHIPLET INTERFACE

    公开(公告)号:US20240004815A1

    公开(公告)日:2024-01-04

    申请号:US17853842

    申请日:2022-06-29

    CPC classification number: G06F13/364 G06F9/4893

    Abstract: Systems and methods are disclosed for scheduling a data link training by a controller. The system and method include receiving an indication that a physical layer of a data link is not transferring data and initiating a training process of the physical layer of the data link in response to the indication that the physical layer of the data link is not transferring data. In one aspect, the indication that the physical layer of a data link is not transferring data is an indication that the physical layer of the data link is in a low power state. In another aspect, the indication that the physical layer of a data link is not transferring data is an indication that a data transfer has been completed.

    Precise shadowing and adjustment of on-die timers in low power states

    公开(公告)号:US11734151B2

    公开(公告)日:2023-08-22

    申请号:US17357200

    申请日:2021-06-24

    CPC classification number: G06F11/349 G06F11/3409

    Abstract: An integrated circuit (IC) includes a first circuit including a timer for receiving an adjustable clock signal. Responsive to leaving the non-operational power state to enter a power state in which the adjustable clock has a lower frequency than the reference clock, the first circuit adjusts the frequency of the adjustable clock to a frequency higher than the lower frequency, and then receives an elapsed time associated with the non-operational power state and starts the timer using an adjusted timer value.

    Demand based probe filter initialization after low power state

    公开(公告)号:US11703932B2

    公开(公告)日:2023-07-18

    申请号:US17357047

    申请日:2021-06-24

    CPC classification number: G06F1/3234

    Abstract: A data fabric routes requests between the plurality of requestors. A probe filter tracks the state of cached lines of memory at a probe filter coupled to the data fabric. Responsive to the data fabric leaving a non-operational power state while all requestors that are probe filter clients are in a non-operational power state, the power management controller delays a probe filter initialization state in which data regarding cached lines is initialized following the non-operational power state.

    MECHANISM FOR PERFORMING DISTRIBUTED POWER MANAGEMENT OF A MULTI-GPU SYSTEM

    公开(公告)号:US20220091657A1

    公开(公告)日:2022-03-24

    申请号:US17031739

    申请日:2020-09-24

    Inventor: Benjamin Tsien

    Abstract: Systems, apparatuses, and methods for efficient power management of a multi-node computing system are disclosed. A computing system includes multiple nodes that receive tasks to process. The nodes include a processor, local memory, a power controller, and multiple link interfaces for transferring messages with other nodes across links. Using a distributed approach for power management, negotiation for powering down components of the computing system occurs without performing a centralized system-wide power down. Each node is able to power down its links, its processor and other components regardless of whether other components of the computing system are still active or powered up. A link interface initiates power down of a link with delay or without delay based on a prediction of whether a link idle condition leads to the link interface remaining idle for at least a target idle threshold period of time.

Patent Agency Ranking