Speculative register reclamation
    1.
    发明授权

    公开(公告)号:US12254318B2

    公开(公告)日:2025-03-18

    申请号:US18143990

    申请日:2023-05-05

    Inventor: Sanyam Mehta

    Abstract: A system determines an original instruction with a first logical register (LR) mapped to a first physical register (PR). The system determines a current instruction with a current LR. A prior instruction is associated with a second LR mapped to a second PR. The system allocates the current LR to a third PR. Responsive to determining that the current and prior instructions are executed in different iterations, the system marks the second PR as not eligible for early release. Responsive to determining that the current LR is previously mapped to the first PR, the allocation comprises a redefinition of the first LR. Responsive to determining that the first PR is eligible for early release and that the current and original instructions are executed in the same or consecutive iterations, the system releases the first PR based upon the redefinition and not the prior instruction completing or the current instruction committing.

    METHOD AND SYSTEM FOR HARDWARE-ASSISTED PRE-EXECUTION

    公开(公告)号:US20230315471A1

    公开(公告)日:2023-10-05

    申请号:US18328099

    申请日:2023-06-02

    Inventor: Sanyam Mehta

    CPC classification number: G06F9/3814 G06F9/30065 G06F9/321 G06F9/5016

    Abstract: One aspect provides a system for hardware-assisted pre-execution. During operation, the system determines a pre-execution code region comprising one or more instructions. The system increments a global counter upon initiating the one or more instructions. The system issues a first instruction, which involves setting, in a first entry for the first instruction in a data structure, a first prefetch region identifier with a current value of the global counter. Responsive to a head pointer of the data structure reaching the first entry, the system: determines, based on a non-zero value for the first prefetch region identifier, that the first entry is not available to be allocated; and advances the head pointer to a next entry in the data structure, which renders a load associated with the first entry as a non-blocking load. The system resets the global counter upon completing the one or more instructions.

    Method and system for hard ware-assisted pre-execution

    公开(公告)号:US11687344B2

    公开(公告)日:2023-06-27

    申请号:US17412200

    申请日:2021-08-25

    Inventor: Sanyam Mehta

    Abstract: One aspect provides a system for hardware-assisted pre-execution. During operation, the system determines a pre-execution code region comprising one or more instructions. The system increments a global counter upon initiating the one or more instructions. The system issues a first instruction, which involves setting, in a first entry for the first instruction in a data structure, a first prefetch region identifier with a current value of the global counter. Responsive to a head pointer of the data structure reaching the first entry, the system: determines, based on a non-zero value for the first prefetch region identifier, that the first entry is not available to be allocated; and advances the head pointer to a next entry in the data structure, which renders a load associated with the first entry as a non-blocking load. The system resets the global counter upon completing the one or more instructions.

    SPECULATIVE REGISTER RECLAMATION
    4.
    发明公开

    公开(公告)号:US20240248719A1

    公开(公告)日:2024-07-25

    申请号:US18143990

    申请日:2023-05-05

    Inventor: Sanyam Mehta

    CPC classification number: G06F9/3013 G06F9/30058 G06F9/3861

    Abstract: A system determines an original instruction with a first logical register (LR) mapped to a first physical register (PR). The system determines a current instruction with a current LR. A prior instruction is associated with a second LR mapped to a second PR. The system allocates the current LR to a third PR. Responsive to determining that the current and prior instructions are executed in different iterations, the system marks the second PR as not eligible for early release. Responsive to determining that the current LR is previously mapped to the first PR, the allocation comprises a redefinition of the first LR. Responsive to determining that the first PR is eligible for early release and that the current and original instructions are executed in the same or consecutive iterations, the system releases the first PR based upon the redefinition and not the prior instruction completing or the current instruction committing.

    METHOD AND SYSTEM FOR HARDWARE-ASSISTED PRE-EXECUTION

    公开(公告)号:US20230061576A1

    公开(公告)日:2023-03-02

    申请号:US17412200

    申请日:2021-08-25

    Inventor: Sanyam Mehta

    Abstract: One aspect provides a system for hardware-assisted pre-execution. During operation, the system determines a pre-execution code region comprising one or more instructions. The system increments a global counter upon initiating the one or more instructions. The system issues a first instruction, which involves setting, in a first entry for the first instruction in a data structure, a first prefetch region identifier with a current value of the global counter. Responsive to a head pointer of the data structure reaching the first entry, the system: determines, based on a non-zero value for the first prefetch region identifier, that the first entry is not available to be allocated; and advances the head pointer to a next entry in the data structure, which renders a load associated with the first entry as a non-blocking load. The system resets the global counter upon completing the one or more instructions.

    Method and system for hardware-assisted pre-execution

    公开(公告)号:US12079631B2

    公开(公告)日:2024-09-03

    申请号:US18328099

    申请日:2023-06-02

    Inventor: Sanyam Mehta

    Abstract: One aspect provides a system for hardware-assisted pre-execution. During operation, the system determines a pre-execution code region comprising one or more instructions. The system increments a global counter upon initiating the one or more instructions. The system issues a first instruction, which involves setting, in a first entry for the first instruction in a data structure, a first prefetch region identifier with a current value of the global counter. Responsive to a head pointer of the data structure reaching the first entry, the system: determines, based on a non-zero value for the first prefetch region identifier, that the first entry is not available to be allocated; and advances the head pointer to a next entry in the data structure, which renders a load associated with the first entry as a non-blocking load. The system resets the global counter upon completing the one or more instructions.

    OPTIMIZING APPLICATION EXECUTION BASED ON MEMORY-LEVEL PARALLELISM (MLP)-BASED METRICS

    公开(公告)号:US20230359358A1

    公开(公告)日:2023-11-09

    申请号:US17662356

    申请日:2022-05-06

    Inventor: Sanyam Mehta

    Abstract: A process includes determining a memory bandwidth of a processor subsystem corresponding to an execution of an application by the processor subsystem. The process includes determining an average memory latency corresponding to the execution of the application and determining an average occupancy of a miss status handling register queue associated with the execution of the application based on the memory bandwidth and the average memory latency. The process includes, based on the average occupancy of the miss status handling register queue and a capacity of the miss status handling register queue, generating data that represents a recommendation of an optimization to be applied to the application.

    REGION-AWARE POWER & ENERGY REGULATION

    公开(公告)号:US20250155955A1

    公开(公告)日:2025-05-15

    申请号:US18388573

    申请日:2023-11-10

    Abstract: A region-aware power/energy regulation method comprises periodically identifying a region of an application which is currently being executed by a given processor out of a plurality of regions of the application. The method also comprises measuring the instructions per second (IPS) of the given processor during execution of the identified region. The method also comprises determining a compute-boundedness parameter of the identified region based on the measured IPS. The method also comprises determining an optimal frequency for the identified region based on the compute-boundedness parameter thereof, and instructing the given processor to set a frequency of the given processor to the optimal frequency during execution of the identified region.

    Method and system for selective early release of physical registers based on a release field value in a scheduler

    公开(公告)号:US11531544B1

    公开(公告)日:2022-12-20

    申请号:US17388666

    申请日:2021-07-29

    Inventor: Sanyam Mehta

    Abstract: The system creates, in a scheduler data structure, a first entry for a consumer instruction associated with a logical register ID. The first entry includes: a scheduler entry ID; a physical register ID allocated for the logical register ID; a checkpoint ID; one or more scheduler entry IDs for one or more prior producer instructions; and a release field which indicates whether to early release a physical register. The system updates a register alias table entry to include the scheduler entry ID and the checkpoint ID of the consumer instruction. The system receives the scheduler entry ID and a checkpoint ID for a respective prior producer instruction. Responsive to determining that the received checkpoint ID does not match the checkpoint ID associated with the consumer instruction, the system sets a release field to indicate that a physical register is to remain allocated.

Patent Agency Ranking