公开(公告)号：US11347994B2

公开(公告)日：2022-05-31

申请号：US16160466

申请日：2018-10-15

Applicant: Intel Corporation

Inventor： Amrita Mathuriya ,  Sasikanth Manipatruni ,  Victor Lee ,  Huseyin Sumbul ,  Gregory Chen ,  Raghavan Kumar ,  Phil Knag ,  Ram Krishnamurthy ,  Ian Young ,  Abhishek Sharma

IPC: G06F3/06 ,  G06N3/04 ,  G06F12/0875 ,  G06N3/063

Abstract: The present disclosure is directed to systems and methods of bit-serial, in-memory, execution of at least an nth layer of a multi-layer neural network in a first on-chip processor memory circuitry portion contemporaneous with prefetching and storing layer weights associated with the (n+1)st layer of the multi-layer neural network in a second on-chip processor memory circuitry portion. The storage of layer weights in on-chip processor memory circuitry beneficially decreases the time required to transfer the layer weights upon execution of the (n+1)st layer of the multi-layer neural network by the first on-chip processor memory circuitry portion. In addition, the on-chip processor memory circuitry may include a third on-chip processor memory circuitry portion used to store intermediate and/or final input/output values associated with one or more layers included in the multi-layer neural network.

公开(公告)号：US11294985B2

公开(公告)日：2022-04-05

申请号：US16175229

申请日：2018-10-30

Applicant: INTEL CORPORATION

Inventor： Amrita Mathuriya ,  Sasikanth Manipatruni ,  Dmitri Nikonov ,  Ian Young ,  Ram Krishnamurthy

IPC: G06F17/16 ,  G11C7/12 ,  G06N3/063

Abstract: Techniques are provided for efficient matrix multiplication using in-memory analog parallel processing, with applications for neural networks and artificial intelligence processors. A methodology implementing the techniques according to an embodiment includes storing two matrices in-memory. The first matrix is stored in transposed form such that the transposed first matrix has the same number of rows as the second matrix. The method further includes reading columns of the matrices from the memory in parallel, using disclosed bit line functional read operations and cross bit line functional read operations, which are employed to generate analog dot products between the columns. Each of the dot products corresponds to an element of the matrix multiplication product of the two matrices. In some embodiments, one of the matrices may be used to store neural network weighting factors, and the other matrix may be used to store input data to be processed by the neural network.

公开(公告)号：US11276730B2

公开(公告)日：2022-03-15

申请号：US16246360

申请日：2019-01-11

Applicant: Intel Corporation

Inventor： Kevin O'Brien ,  Christopher Wiegand ,  Tofizur Rahman ,  Noriyuki Sato ,  Gary Allen ,  James Pellegren ,  Angeline Smith ,  Tanay Gosavi ,  Sasikanth Manipatruni ,  Kaan Oguz ,  Benjamin Buford ,  Ian Young

IPC: H01L27/22 ,  H01L43/04 ,  H01L43/14 ,  H01L43/08 ,  H01L43/10

Abstract: A perpendicular spin orbit memory device includes a first electrode having a magnetic material and platinum and a material layer stack on a portion of the first electrode. The material layer stack includes a free magnet, a fixed magnet above the first electrode, a tunnel barrier between the free magnet and the fixed magnet and a second electrode coupled with the fixed magnet.

公开(公告)号：US20220069009A1

公开(公告)日：2022-03-03

申请号：US17399530

申请日：2021-08-11

Applicant: Intel Corporation

Inventor： Sasikanth Manipatruni ,  Dmitri E. Nikonov ,  Ian A. Young

IPC: H01L27/22 ,  H01L43/12 ,  H01L43/10 ,  H01L43/02

Abstract: A three dimensional (3D) array of magnetic random access memory (MRAM) bit-cells is described, wherein the array includes a mesh of: a first interconnect extending along a first axis; a second interconnect extending along a second axis; and a third interconnect extending along a third axis, wherein the first, second and third axes are orthogonal to one another, and wherein a bit-cell of the MRAM bit-cells includes: a magnetic junction device including a first electrode coupled to the first interconnect; a piezoelectric (PZe) layer adjacent to a second electrode, wherein the second electrode is coupled to the second interconnect; and a first layer adjacent to the PZe layer and the magnetic junction, wherein the first layer is coupled the third interconnect.

公开(公告)号：US11139389B2

公开(公告)日：2021-10-05

申请号：US15427968

申请日：2017-02-08

Applicant: Intel Corporation

Inventor： Sasikanth Manipatruni ,  Dmitri E. Nikonov ,  Ian A Young

IPC: H01L29/66 ,  H03K19/16 ,  G11C11/16 ,  H01L27/22 ,  H01L43/08 ,  H01L43/10 ,  H01L43/12

Abstract: Described is an apparatus, for spin state element device, which comprises: a variable resistive magnetic (VRM) device to receive a magnetic control signal to adjust resistance of the VRM device; and a magnetic logic gating (MLG) device, coupled to the VRM device, to receive a magnetic logic input and perform logic operation on the magnetic logic input and to drive an output magnetic signal based on the resistance of the VRM device. Described is a magnetic de-multiplexer which comprises: a first VRM device to receive a magnetic control signal to adjust resistance of the first VRM; a second VRM device to receive the magnetic control signal to adjust resistance of the second VRM device; and an MLG device, coupled to the first and second VRM devices, the MLG device having at least two output magnets to output magnetic signals based on the resistances of the first and second VRM devices.

公开(公告)号：US11069609B2

公开(公告)日：2021-07-20

申请号：US16647691

申请日：2017-11-03

Applicant: INTEL CORPORATION

Inventor： Sasikanth Manipatruni ,  Jasmeet S. Chawla ,  Chia-Ching Lin ,  Dmitri E. Nikonov ,  Ian A. Young ,  Robert L. Bristol

IPC: H01L23/522 ,  H01F10/32 ,  H01L21/768 ,  H01L23/528 ,  H01L27/22 ,  H01L43/02

Abstract: Techniques are disclosed for forming vias for integrated circuit structures. During an additive via formation process, a dielectric material is deposited, an etch stop layer is deposited, a checkerboard pattern is deposited on the etch stop layer, regions in the checkerboard pattern are removed where it is desired to have vias, openings are etched in the dielectric material through the removed regions, and the openings are filled with a first via material. This is then repeated for a second via material. During the subtractive via formation process, a first via material is deposited, an etch stop layer is deposited, a checkerboard pattern is deposited on the etch stop layer, regions in the checkerboard pattern are removed where it is not desired to have vias, openings are etched in the first via material through the removed regions. This is then repeated for a second via material.

公开(公告)号：US11016701B2

公开(公告)日：2021-05-25

申请号：US16146878

申请日：2018-09-28

Applicant: Intel Corporation

Inventor： Ian Young ,  Ram Krishnamurthy ,  Sasikanth Manipatruni ,  Amrita Mathuriya ,  Abhishek Sharma ,  Raghavan Kumar ,  Phil Knag ,  Huseyin Sumbul ,  Gregory Chen

IPC: G06F3/06 ,  G11C7/10 ,  H03K3/03

Abstract: Techniques and mechanisms for a memory device to perform in-memory computing based on a logic state which is detected with a voltage-controlled oscillator (VCO). In an embodiment, a VCO circuit of the memory device receives from a memory array a first signal indicating a logic state that is based on one or more currently stored data bits. The VCO provides a conversion from the logic state being indicated by a voltage characteristic of the first signal to the logic state being indicated by a corresponding frequency characteristic of a cyclical signal. Based on the frequency characteristic, the logic state is identified and communicated for use in an in-memory computation at the memory device. In another embodiment, a result of the in-memory computation is written back to the memory array.

公开(公告)号：US10884957B2

公开(公告)日：2021-01-05

申请号：US16160952

申请日：2018-10-15

Applicant: Intel Corporation

Inventor： Amrita Mathuriya ,  Sasikanth Manipatruni ,  Victor W. Lee ,  Abhishek Sharma ,  Huseyin E. Sumbul ,  Gregory Chen ,  Raghavan Kumar ,  Phil Knag ,  Ram Krishnamurthy ,  Ian Young

IPC: G06F13/16 ,  G06F13/28 ,  G06N20/00

Abstract: Techniques and mechanisms for performing in-memory computations with circuitry having a pipeline architecture. In an embodiment, various stages of a pipeline each include a respective input interface and a respective output interface, distinct from said input interface, to couple to different respective circuitry. These stages each further include a respective array of memory cells and circuitry to perform operations based on data stored by said array. A result of one such in-memory computation may be communicated from one pipeline stage to a respective next pipeline stage for use in further in-memory computations. Control circuitry, interconnect circuitry, configuration circuitry or other logic of the pipeline precludes operation of the pipeline as a monolithic, general-purpose memory device. In other embodiments, stages of the pipeline each provide a different respective layer of a neural network.

公开(公告)号：US20200343301A1

公开(公告)日：2020-10-29

申请号：US16396451

申请日：2019-04-26

Applicant: Intel Corporation

Inventor： Benjamin Buford ,  Angeline Smith ,  Noriyuki Sato ,  Tanay Gosavi ,  Kaan Oguz ,  Christopher Wiegand ,  Kevin O'Brien ,  Tofizur Rahman ,  Gary Allen ,  Sasikanth Manipatruni ,  Emily Walker

IPC: H01L27/22 ,  H01L43/14 ,  H01L43/08 ,  H01L43/06 ,  H01L43/04

Abstract: A memory apparatus includes a first electrode having a spin orbit material. The memory apparatus further includes a first memory device on a portion of the first electrode and a first dielectric adjacent to a sidewall of the first memory device. The memory apparatus further includes a second memory device on a portion of the first electrode and a second dielectric adjacent to a sidewall of the second memory device. A second electrode is on and in contact with a portion of the first electrode, where the second electrode is between the first memory device and the second memory device. The second electrode has a lower electrical resistance than an electrical resistance of the first electrode. The memory apparatus further includes a first interconnect structure and a second interconnect, each coupled with the first electrode.

公开(公告)号：US10748603B2

公开(公告)日：2020-08-18

申请号：US16146473

申请日：2018-09-28

Applicant: Intel Corporation

Inventor： Huseyin Ekin Sumbul ,  Gregory K. Chen ,  Raghavan Kumar ,  Phil Knag ,  Abhishek Sharma ,  Sasikanth Manipatruni ,  Amrita Mathuriya ,  Ram Krishnamurthy ,  Ian A. Young

IPC: G11C8/00 ,  G11C11/418 ,  G06F9/30 ,  G11C11/419 ,  G11C13/00 ,  G11C7/10 ,  G11C11/54 ,  G06N3/063 ,  G06N3/08 ,  G11C7/18 ,  G06F7/544 ,  G11C11/16

Abstract: A memory circuit has compute-in-memory circuitry that enables a multiply-accumulate (MAC) operation based on shared charge. Row access circuitry drives multiple rows of a memory array to multiply a first data word with a second data word stored in the memory array. The row access circuitry drives the multiple rows based on the bit pattern of the first data word. Column access circuitry drives a column of the memory array when the rows are driven. Accessed rows discharge the column line in an accumulative fashion. Sensing circuitry can sense voltage on the column line. A processor in the memory circuit computes a MAC value based on the voltage sensed on the column.