OPTIMAL PLACEMENT OF DATA STRUCTURES IN A HYBRID MEMORY BASED INFERENCE COMPUTING PLATFORM

    公开(公告)号:US20210357138A1

    公开(公告)日:2021-11-18

    申请号:US15929618

    申请日:2020-05-13

    Abstract: In a deep neural network (DNN), weights are defined that represent a strength of connections between different neurons of the DNN and activations are defined that represent an output produced by a neuron after passing through an activation function of receiving an input and producing an output based on some threshold value. The weight traffic associated with a hybrid memory therefore is distinguished from the activation traffic to the hybrid memory, and one or more data structures may be dynamically allocated in the hybrid memory according to the weights and activations of the or more data structures in the DNN. The hybrid memory includes at least a first memory and a second memory that differ according to write endurance attributes.

    TERNARY CONTENT ADDRESSABLE MEMORY
    10.
    发明申请
    TERNARY CONTENT ADDRESSABLE MEMORY 审中-公开
    三进制内容可寻址内存

    公开(公告)号:US20170040059A1

    公开(公告)日:2017-02-09

    申请号:US14818764

    申请日:2015-08-05

    CPC classification number: G11C15/04

    Abstract: Ternary content addressable memory (TCAM) structures and methods of use are disclosed. The memory architecture includes one or more ternary content addressable memory (TCAM) fields, and control logic that applies progressively discriminating data-masking and scores a closeness of a match based on matched and mismatched bits.

    Abstract translation: 公开了三元内容可寻址存储器(TCAM)结构和使用方法。 存储器架构包括一个或多个三元内容可寻址存储器(TCAM)字段,以及控制逻辑,其应用逐行识别数据掩蔽,并且基于匹配和不匹配的比特对匹配的接近进行评分。

Patent Agency Ranking