Usage-Based Disturbance Mitigation
    2.
    发明公开

    公开(公告)号:US20240347098A1

    公开(公告)日:2024-10-17

    申请号:US18624920

    申请日:2024-04-02

    发明人: Yang Lu

    IPC分类号: G11C11/406 G11C11/4093

    摘要: Apparatuses and techniques for implementing usage-based disturbance mitigation are described. In some examples, a Total Mitigation Pump Pair (TMPP) queue and a dynamic Mitigation Threshold (MT) can improve the mitigation efficiency of usage-based disturbances while reducing power requirements and increasing performance impact. In various aspects, the TMPP queue is a first-in first out (FIFO) queue useful to implement the described disturbance mitigation techniques.

    Receiver circuit, memory device and operation method using the same

    公开(公告)号:US12106798B2

    公开(公告)日:2024-10-01

    申请号:US18358049

    申请日:2023-07-25

    发明人: Chi-Sing Lo

    IPC分类号: G11C11/4093

    CPC分类号: G11C11/4093

    摘要: A receiver circuit that includes a pair of pre-stage amplifier circuits and a post-stage amplifier circuit is introduced. The pre-stage amplifier circuits are configured to receive an input signal and a reference voltage signal, output first pre-stage amplifying signals through a first connection node and a second connection node separately, and output second pre-stage amplifying signals through a third connection node and a fourth connection node separately. The post-stage amplifier circuit is configured to receive the first pre-stage amplifying signals and the second pre-stage amplifying signals from the pair of pre-stage amplifier circuits through the first connection node, the second connection node, the third connection node and the fourth connection node separately, and output a post amplifying signal according to the first pre-stage amplifying signals and the second pre-stage amplifying signals.

    TECHNIQUES AND DEVICES TO REDUCE BUS CROSS TALK FOR MEMORY SYSTEMS

    公开(公告)号:US20240321340A1

    公开(公告)日:2024-09-26

    申请号:US18588686

    申请日:2024-02-27

    摘要: Methods, systems, and devices for techniques and devices to reduce bus cross talk are described. Adjacent conductive lines in a bus of a memory system may be electrically coupled if both conductive lines are concurrently driven to a high-state. For example, the bus may include a logic circuit coupled between the adjacent conductive lines, which may selectively couple the conductive lines based on the voltage applied to each conductive line. In some examples, the logic circuit may include an input coupled to the control signals of the drivers associated with the adjacent conductive lines. If both control signals are concurrently high, the logic circuit may activate a transistor to couple the conductive lines. Such electrical coupling may reduce or eliminate the capacitive coupling between the two conductive lines when both are driven to a high state, which may result in increased reliability of signals in the conductive lines.

    Multichannel memory to augment local memory

    公开(公告)号:US12094525B2

    公开(公告)日:2024-09-17

    申请号:US17814254

    申请日:2022-07-22

    摘要: A memory system, a method of assembling the memory system, and a computer system. The memory system includes a global memory device coupled to a plurality of processing elements. The global memory device is positioned external to a chip on which the plurality of processing devices reside. The memory system also includes at least one main scratchpad coupled to the at least one processing element of the plurality of processing devices and the global memory device. The memory system further includes a plurality of auxiliary scratchpads coupled to the plurality of processing elements and the global memory device. The one or more auxiliary scratchpads are configured to store static tensors. At least a portion of the plurality of auxiliary scratchpads are configured as a unitary multichannel device.

    ADDRESS MAPPING FOR IMPROVED RELIABILITY
    10.
    发明公开

    公开(公告)号:US20240274181A1

    公开(公告)日:2024-08-15

    申请号:US18584669

    申请日:2024-02-22

    申请人: Rambus Inc.

    摘要: The internal row addressing of each DRAM on a module is mapped such that row hammer affects different neighboring row addresses in each DRAM. Because the external row address to internal row address mapping scheme ensures that each set of neighboring rows for a given externally addressed row is different for each DRAM on the module, row hammering of a given externally addressed row spreads the row hammering errors across different externally addressed rows on each DRAM. This has the effect of confining the row hammer errors for each row that is hammered to a single DRAM per externally addressed neighboring row. By confining the row hammer errors to a single DRAM, the row hammer errors are correctible using a single device data correct (SDDC) scheme.