NEURAL PROCESSOR
    3.
    发明申请
    NEURAL PROCESSOR 审中-公开

    公开(公告)号:US20200026978A1

    公开(公告)日:2020-01-23

    申请号:US16552619

    申请日:2019-08-27

    Abstract: A neural processor. In some embodiments, the processor includes a first tile, a second tile, a memory, and a bus. The bus may be connected to the memory, the first tile, and the second tile. The first tile may include: a first weight register, a second weight register, an activations buffer, a first multiplier, and a second multiplier. The activations buffer may be configured to include: a first queue connected to the first multiplier and a second queue connected to the second multiplier. The first queue may include a first register and a second register adjacent to the first register, the first register being an output register of the first queue. The first tile may be configured: in a first state: to multiply, in the first multiplier, a first weight by an activation from the output register of the first queue, and in a second state: to multiply, in the first multiplier, the first weight by an activation from the second register of the first queue.

    NEURAL PROCESSOR
    7.
    发明申请
    NEURAL PROCESSOR 审中-公开

    公开(公告)号:US20200026980A1

    公开(公告)日:2020-01-23

    申请号:US16552945

    申请日:2019-08-27

    Abstract: A neural processor. In some embodiments, the processor includes a first tile, a second tile, a memory, and a bus. The bus may be connected to the memory, the first tile, and the second tile. The first tile may include: a first weight register, a second weight register, an activations buffer, a first multiplier, and a second multiplier. The activations buffer may be configured to include: a first queue connected to the first multiplier and a second queue connected to the second multiplier. The first queue may include a first register and a second register adjacent to the first register, the first register being an output register of the first queue. The first tile may be configured: in a first state: to multiply, in the first multiplier, a first weight by an activation from the output register of the first queue, and in a second state: to multiply, in the first multiplier, the first weight by an activation from the second register of the first queue.

    NEURAL PROCESSOR
    8.
    发明申请
    NEURAL PROCESSOR 审中-公开

    公开(公告)号:US20190392287A1

    公开(公告)日:2019-12-26

    申请号:US16446610

    申请日:2019-06-19

    Abstract: A neural processor. In some embodiments, the processor includes a first tile, a second tile, a memory, and a bus. The bus may be connected to the memory, the first tile, and the second tile. The first tile may include: a first weight register, a second weight register, an activations buffer, a first multiplier, and a second multiplier. The activations buffer may be configured to include: a first queue connected to the first multiplier and a second queue connected to the second multiplier. The first queue may include a first register and a second register adjacent to the first register, the first register being an output register of the first queue. The first tile may be configured: in a first state: to multiply, in the first multiplier, a first weight by an activation from the output register of the first queue, and in a second state: to multiply, in the first multiplier, the first weight by an activation from the second register of the first queue.

    Method performed by a base station, base station and computer-readable storage medium

    公开(公告)号:US12255721B2

    公开(公告)日:2025-03-18

    申请号:US18313568

    申请日:2023-05-08

    Abstract: A method which may be performed by a base station, may include obtaining a predicted transmission configuration indicator (TCI) subset for a user equipment (UE); and determining the TCI subset corresponding to the UE based on the predicted TCI subset. The step of obtaining the predicted TCI subset in the solution may be implemented by using a trained artificial intelligence model. Since the beams in the predicted TCI subset may fit the movement trend of the UE, then the target serving beam for the UE when performing serving beam switching may belong to the predicted TCI subset with a greater possibility, thus reducing the occurrence of high delay scenarios when the UE performs the serving beam switching, and thus reducing the beam switching delay.

Patent Agency Ranking