DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claims 1-28 are pending.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 04/24/2024 and 09/09/2024 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 1-28 are rejected under 35 U.S.C. 103 as being unpatentable over Sengupta et al. (Sengupta), NPL “Encoding neural and synaptic functionalities in electron spin: A pathway to efficient neuromorphic computing”, published on December 26, 2017, 24 pages, and further in view of Woo et al., NPL “Enhanced spin-orbit torques in Pt/Co/Ta heterostructures”, published on November 25, 2014 (Both Sengupta and Woo are from IDS filed on 09/09/2024).
As to independent claim 1, Sengupta discloses a deep neural network (DNN) device (page 6, left column, second paragraph and Figure 10: “Such neurons offer high recognition accuracies in a vast category of large-scale recognition tasks and are routinely utilized today as a basic building block of deep neural networks), the DNN device comprising:
one or more first devices configured to perform a first operation of a neural node of a deep neural network (DNN) (page 5, right column, fourth paragraph: signal being transmitted from the pre-neuron”), the one or more first devices each comprising:
a first spin orbit torque (SOT1) layer (Figure 2b on page3: interface between heavy mental HM and ferromagnet FL; page 4, right column, second paragraph: when a charge current flows through the underlying HM, spin-orbit torque (SOT1) is generated at the FM-HM interface);
a ferromagnetic layer disposed between the SOT1 (Figure. 2b: ferromagnet FL);
a first terminal coupled to the SOT1 layer (Figure 6: terminal T2 in topmost figure 6);
a second terminal coupled over the stack layers (Figure 6: terminal T1 in topmost figure 6), the second terminal being configured for an input current path that is perpendicular to a plane of the stack layers, the input current path being configured to extend into the ferromagnetic layer (Figure 6: see topmost of Figure 6; page 12, right column, third paragraph: terminal 1…enables the ‘read’ current path in the device between terminals T1 and T3”); and
a third terminal coupled to the stack layers (Figure 6: terminal T3 in topmost figure of Figure 6), the third terminal being configured for an output current path that is in plane of the stack layers (Figures 6 and 10: as it is shown that the terminal t3 is the end terminal of read current and connected to subsequent or post neuron; also shown explicitly from the direction of the read current depicted in Figure 4a of reference 6 of D1 which discloses a similar device, see the arrows indicating the read current flowing through T1 and T3 (from terminal T1 to terminal T3)).
Sengupta, however, does not disclose a second spin orbit torque (SOT2) layer above the ferromagnetic layer. However, Sengupta discloses the need for improving the efficiency of the device (current induced magnetization switching efficiency) through various techniques one of which is through innovations in the material stack (page 6, first paragraph of section E). In the same field of endeavor, Woo discloses of increasing this efficiency by sandwiching the ferromagnet between two SOT layers, which is achieved by having both the bottom and the top surfaces of the ferromagnet to be interfaced with heavy metals (Abstract, first paragraph of left column of page 1 and Figure 1(a)).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the system of Sengupta to include “a second spin orbit torque (SOT2) layer above the ferromagnetic layer”, as taught by Woo for the purpose of improving the efficiency of the device.
As to independent claim 11, Sengupta discloses a deep neural network (DNN) device (page 6, left column, second paragraph and Figure 10: “Such neurons offer high recognition accuracies in a vast category of large-scale recognition tasks and are routinely utilized today as a basic building block of deep neural networks), the DNN device comprising:
a plurality of rows (Figures 3c and 10a);
a plurality of columns (Figures 3c and 10a);
a spin-orbit spin-orbit (SO-SO) device disposed at each cross-point between rows of the plurality of rows and columns of the plurality of columns configured to perform a first operation of a neural node of a deep neural network (DNN) (Figure 10 and page 17, left column, second paragraph: “the computation can be represented by a crossbar array of dimension M x n (Figure 10a). At each cross-point of the array, a spintronic synaptic device is present”), each SO-SO device comprising:
a first spin orbit torque (SOT1) layer (Figure 2b on page3: interface between heavy mental HM and ferromagnet FL; page 4, right column, second paragraph: when a charge current flows through the underlying HM, spin-orbit torque (SOT1) is generated at the FM-HM interface);
a ferromagnetic layer disposed above the SOT1 (Figure. 2b: ferromagnet FL),
wherein a plurality of input currents are configured to be applied to each SO-SO device disposed in each row of the plurality of rows (Figure 3b, page 5, right column, last paragraph: each neural computing unit receives a set of inputs from other pre-neurons through synaptic junctions. The weighted contribution from all the neurons is then summed up and processed by the neurons); and
a plurality of output SO-SO devices disposed within each column configured to perform a second operation of a neural node of a deep neural network (DNN) (Figure 10: post neurons), each output SO-SO device comprising:
a SOT1 layer (Figure 2b on page 3);
a ferromagnetic layer disposed above the SOT1 layer (Figure 2b: ferromagnet FL), wherein an output current of each SO-SO device disposed in each column of the plurality of columns is configured to be input as an input current into the plurality of output SO-SO devices (Figure 10b: pre-neurons; Figures 3a and 10: disposed in each column of the plurality of columns is configured to be input as an input current into the plurality of output SO-SO devices)).
Sengupta, however, does not disclose a second spin orbit torque (SOT2) layer above the ferromagnetic layer. However, Sengupta discloses the need for improving the efficiency of the device (current induced magnetization switching efficiency) through various techniques one of which is through innovations in the material stack (page 6, first paragraph of section E). In the same field of endeavor, Woo discloses of increasing this efficiency by sandwiching the ferromagnet between two SOT layers, which is achieved by having both the bottom and the top surfaces of the ferromagnet to be interfaced with heavy metals (Abstract, first paragraph of left column of page 1 and Figure 1(a)).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the system of Sengupta to include “a second spin orbit torque (SOT2) layer above the ferromagnetic layer”, as taught by Woo for the purpose of improving the efficiency of the device.
As to independent claim 20, Sengupta discloses A deep neural network (DNN) device (page 6, left column, second paragraph and Figure 10: “Such neurons offer high recognition accuracies in a vast category of large-scale recognition tasks and are routinely utilized today as a basic building block of deep neural networks), the DNN device comprising:
a first layer (page 9, left column, first paragraph: neurons are arranged in different layers) of a plurality of first spin orbit torque (SOT) devices (Figure 2b) configured to perform a first operation of a neural node of a deep neural network (DNN) (Figure 3b, page 5, right column, last paragraph: a set of neurons in a particular layer receiving a set of inputs through synaptic weights, the computation can be mapped to a parallel dot-product operation between the inputs and synaptic weights followed by neural processing for each neuron in the layer), each first SOT device comprising:
a first spin orbit torque layer (Figure 2b on page3: interface between heavy mental HM and ferromagnet FL; page 4, right column, second paragraph: when a charge current flows through the underlying HM, spin-orbit torque (SOT1) is generated at the FM-HM interface);
a second layer coupled to a first output terminal (Figure 6, page 12, right column, third paragraph: terminal T3 coupled to the PL layer in Figure 6 for read current between terminals T1 and T3); and
a ferromagnetic layer disposed between the first spin orbit torque layer and the second layer (Figure 6: the FL layer), wherein each first SOT device is individually configured to generate, via the second layer, an output current responsive to a direction of the magnetization of the ferromagnetic layer and an input current (Figure 4a of reference 6, Figure 6, page 12, right column, third paragraph: the read current from terminal T1 to terminal T3 of Figure 6; page 12, right column last paragraph: as the magnitude of the input current I sub in increases, the resistance of the neural device reduces due to decrease in the proportion of the AP domain in the MTJ device. This, in turn, causes the current provided by the output transistor (I sub out) to increase”), the output current being output through the first output terminal of each first SOT device (terminal T3); and
a second layer of a plurality of second SOT devices configured to perform a second operation of a neural node of a deep neural network (DNN) (page 5, right column, fourth paragraph: neuromorphic computation relies on … neuro-transmitter release at the synapse due to the incoming action potential from the pre-neuron and the neuroscience mechanisms occurring in the post-neuron to generate an outgoing signal to the next layer of neurons), each second SOT device comprising:
a first spin orbit torque layer coupled to a first input terminal (Figure 2b on page 3: interface between heavy mental HM and ferromagnet FL; Figure 6: terminal T2);
a second layer (Figure 6: the PL layer); and
a ferromagnetic layer disposed between the first spin orbit torque layer and the second layer (Figure 6: the FL layer), wherein the output current of each first SOT device is configured to be input into the first input terminal of each second SOT device (page 14, right column, first paragraph: terminals T1 and T3 of the device are activated and spike voltage signals can be transmitted from the pre-neuron signal through the MTL conductance to provide an equivalent amount of synaptic current to the post-neuron circuit connected to terminal T3).
Sengupta, however, does not disclose a second spin orbit torque (SOT2) layer above the ferromagnetic layer. However, Sengupta discloses the need for improving the efficiency of the device (current induced magnetization switching efficiency) through various techniques one of which is through innovations in the material stack (page 6, first paragraph of section E). In the same field of endeavor, Woo discloses of increasing this efficiency by sandwiching the ferromagnet between two SOT layers, which is achieved by having both the bottom and the top surfaces of the ferromagnet to be interfaced with heavy metals (Abstract, first paragraph of left column of page 1 and Figure 1(a)).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the system of Sengupta to include “a second spin orbit torque (SOT2) layer above the ferromagnetic layer”, as taught by Woo for the purpose of improving the efficiency of the device.
As to dependent claims 2, 14 and 24, Sengupta discloses wherein the SOT1 layer is configured to affect a direction of a magnetization of the ferromagnetic layer based on a weight current received at the first terminal (page 4, section C: the magnitude of injected spin current density is proportional to the magnitude of input charge current density” & “SOT induced magnetization reversal is only possible for IMA magnets … Energy efficient SOT induced domain wall motion has been also observed in FM_HM bilayers…”).
As to dependent claims 3 and 15, Sengupta discloses wherein each of the one or more first devices is further configured to receive an input current at the input current path, and output an output based on the affected direction of magnetization of the ferromagnetic layer (page 12, right column, last paragraph: as the magnitude of the input current increases, the resistance of the neuronal device reduces due to decrease in the proportion of the AP domain in the MTJ device. This, in turn, causes the current provided by the output transistor to increase).
As to dependent claims 4-5, the subject matter of the claims refers to second devices similar to first devices that were described in claim 1, wherein the output current of the first devices is input as input current to the second devices and is disclosed in relation to Figure 10b of Sengupta, where first devices are indicated in Sengupta as pre-neurons and second devices as post-neurons.
As to dependent claims 6 and 22, Sengupta discloses wherein the output current of the one or more first devices are summed together prior to being input into the one or more second devices (page 5, right column, last paragraph: each neural computing unit receives a set of inputs from other pre-neurons through synaptic junctions. The weighted contribution from all the neurons is then summed up and processed the neurons; page 6, left column, The current flowing through a particular memristive synapse is scaled by the device conductance (synaptic scaling operation) and all such
currents gets summed up along the column of the array, according to Kirchhoff’s law, and passes as the resultant input to the neuron).
As to dependent claims 7, 12, and 21, Sengupta discloses wherein each of the first devices further comprises an oxide layer disposed between the ferromagnetic layer and the SOT2 layer (See Figure 2a and 2b: tunneling oxide).
As to dependent claims 8-9, 13, 18, 26 and 28, the subject matter of the claim refers to composition of the SOT layers and the output SO-SO device, and is merely
related to implementation details or one of several options a skilled person would
choose to employ under appropriate circumstances, and thus it lacks an inventive
step (especially as reference 54 of Sengupta does disclose of getting high efficiency
through doping and BiSb composition (see title and second paragraph on the right
column of page 1.)
As to dependent claim 10, Sengupta discloses one or more interlayer between the SOT1 layer and the above layers, the interlayer comprising one or more of MgO
(see the tunneling oxide in Fig. 2a & 2b and note the next excerpt from page 12 second paragraph of first column: "Additionally, in order to increase spin-current injection at the magnet-channel interface, a tunneling barrier like MgO is utilized between the two materials").
As to dependent claim 16, Sengupta discloses that each SO-SO device is further configured to generate the output current responsive to a direction of the magnetization of the ferromagnetic layer and the input current (page 12 right column last paragraph: "As the magnitude of the input current increases, the resistance of the neuronal device reduces due to decrease in the proportion of the AP domain in the MTJ device. This, in turn, causes the current provided by the output transistor to increase",).
Sengupta, however, does not disclose a second spin orbit torque (SOT2) layer above the ferromagnetic layer. However, Sengupta discloses the need for improving the efficiency of the device (current induced magnetization switching efficiency) through various techniques one of which is through innovations in the material stack (page 6, first paragraph of section E). In the same field of endeavor, Woo discloses of increasing this efficiency by sandwiching the ferromagnet between two SOT layers, which is achieved by having both the bottom and the top surfaces of the ferromagnet to be interfaced with heavy metals (Abstract, first paragraph of left column of page 1 and Figure 1(a)).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the system of Sengupta to include “a second spin orbit torque (SOT2) layer above the ferromagnetic layer”, as taught by Woo for the purpose of improving the efficiency of the device.
As to dependent claim 23, Sengupta discloses that each first SOT device further
comprises a second input terminal (terminal T1 Fig. 6) configured to receive the input current, and wherein the input current is directed perpendicularly into the second layer (shown explicitly from the direction of the read current depicted in Fig. 4a of reference 6).
Conclusion
Any inquiry concerning this communication should be directed to CHAU T NGUYEN at telephone number (571)272-4092. The examiner can normally be reached on M-F from 8am to 5pm (PT).
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) Form at https://www.uspto.gov/patents/uspto-automated-interview-request-air-form.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Cesar Paula, can be reached at telephone number 5712724128. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from Patent Center and the Private Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from Patent Center or Private PAIR. Status information for unpublished applications is available through Patent Center and Private PAIR for authorized users only. Should you have questions about access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free).
/CHAU T NGUYEN/Primary Examiner, Art Unit 2145
