Detailed Action
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
This is in response to RCE and amendment filed on 12/11/2025. Claims 1, 13 and 18 are amended, claims 1-21 are pending for examination.
Specification
The lengthy specification has not been checked to the extent necessary to determine the presence of all possible minor errors. Applicant's cooperation is requested in correcting any errors of which applicant may become aware in the specification.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 11/12/2025 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Response to applicant’s Arguments
Applicant’s arguments, see page 7-8, filed 12/11/2025, with respect to the rejection(s) of claims 1, 13 and 18 under 103 rejection have been fully considered and are persuasive. Therefore, the rejection has been withdrawn. However, upon further consideration, a new ground(s) of rejection is made in view of Jacobs et al., US 2023/0244990 A1.
Claim Rejections - 35 USC § 103
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.
Claim 1, 4, 8-20 are rejected under 35 U.S.C. 103 as being unpatentable over AHMED et al., US 2022/0179871 A1, in view of Sims., US 10,565234 B1, further in view of Jacobs et al., US 2023/0244990 A1.
Regarding claim 1, AHMED teaches a system (system 100 in Fig.1 and Fig.5; FIG. 5 is a block diagram illustrating physical components (e.g., hardware) of a computing device 500 with which aspects of the disclosure may be practiced. The computing device components described below may be suitable for the computing devices described above, including devices 102, 104, and 106 in FIG. 1) comprising:
a processor (Fig.5; the combination of the processing unit 502 and a system memory 504; the components within a dashed line 508 as shown in Fig.5), the processor including a local memory (system memory 504; the system memory 504 may comprise, but is not limited to, volatile storage (e.g., random access memory), non-volatile storage (e.g., read-only memory), flash memory, or any combination of such memories), the local memory internal to the processor (Fig.5);
a memory (Fig.5; removable storage 509) connected to the processor (Fig.5);
a cache-coherent interconnect storage device (it is taught as non-removable storage 510) connected to the processor (Fig.5); and
an embedded management unit (EMU) configured to manage the storage of a document embedding vector in the local memory, the memory, or the cache-coherent interconnect storage device (Fig.5; it is taught as the operating system 505, for example, may be suitable for controlling the operation of the computing device 500;
section 0031; query processor 116 generates an embedding vector for a received query. As noted above, data store 114 may store embedding vectors for documents, such that query processor 116 uses the trained model to process the query embedding vector in view of the document embedding vectors stored by data store 114, thereby generating a responsive set of documents).
AHMED further teaches obtaining a request comprising a search query, generating a query embedding vector for the search query; generating a set of documents responsive to the search query based on the query embedding vector and document embedding vectors for documents of a document corpus, providing generated set of documents that is responsive to the search query.
AHMED does not clearly teach a query is received from a host and send a document to the host.
However, Sims teaches a query is received from a host and send a document to the host (col.20, lines 12-18; The document identification process 600 may be invoked in various contexts. The invocation context will determine where the input query is received from and where the response is returned. Typically, however, the input query will be received from either a user device 110 (as configured by a client application 112) or a collaboration system 120 (as configured by a server application 122)).
It would have been obvious to the ordinary skill in the art before the effective filing date of the claimed invention to utilize the teachings Sims into AHMED such as a query is received from a host and send a document to the host because this will identify documents that relate to an input query (col.2, lines 23-24) and facilitate the storage and retrieval of data (col.4, lines 41-43).
AHMED and Sims do not clearly teach the EMU tracks where the document embedding vector is stored in the local memory, the memory, or the cache-coherent interconnect storage device.
However, Jacobs teaches the EMU tracks where the document embedding vector is stored in the local memory, the memory, or the cache-coherent interconnect storage device (section 0013; 0035; 0043 and section 0073; the document embedding vectors computed by the document embedding component can be cached and be re-used; The document embedding component 140 can be configured to compute, for each document, an n-dimensional vector of real numbers. The n-dimensional vector can be herein referred to as the document embedding vector; the computed n-dimensional vectors of the documents can also be cached).
It would have been obvious to the ordinary skill in the art before the effective filing date of the claimed invention to utilize the teachings Jacobs into AHMED and Sims such as the EMU tracks where the document embedding vector is stored in the local memory, the memory, or the cache-coherent interconnect storage device because it efficiently compute approximate embedding vectors of newly added documents. Thus, new documents can be accommodated in the system without the need to immediately re-train the system (section 0013 of Jacobs).
Regarding claim 4, AHMED and Sims teach the claimed invention as shown above, AHMED further teaches the EMU is configured to copy the document embedding vector into the local memory based at least in part on the query (section 0043; the body is obtained from a data source, such as document store 118 of data source 104. As another example, the document may be from a data store (e.g., data store 114 in FIG. 1). For example, the document body may have been cached from the data source in the data store).
Regarding claims 8 and 17, AHMED teaches the EMU is configured to prefetch a second document embedding vector from the cache-coherent interconnect storage device into the memory based at least in part on a query associated with the query embedding vector (section 0050; the document vectors may have been pre-generated (e.g., by a document vectorizer, such as document vectorizer 110 in FIG. 1).; section 0043; the document body may have been cached from the data source in the data store).
Regarding claim 9, AHMED teaches further comprising an accelerator including the processor (section 0023; Server device 102 may process the search query (e.g., using query processor 116) to determine a set of documents that is responsive to the search query).
Regarding claim 10, AHMED teaches the processor is configured to generate a query embedding vector based at least in part on a query and to process the query embedding vector and the document embedding vector (section 0031; query processor 116 uses the trained model to process queries (e.g., as may be received from client application 120 of client device 106) and generate a set of documents that is responsive to the query accordingly. As an example, query processor 116 generates an embedding vector for a received query).
Regarding claim 11, AHMED teaches the processor is configured to perform a similarity search using the query embedding vector and the document embedding vector to generate a result (section 0031; query processor 116 generates an embedding vector for a received query. As noted above, data store 114 may store embedding vectors for documents, such that query processor 116 uses the trained model to process the query embedding vector in view of the document embedding vectors stored by data store 114, thereby generating a responsive set of documents).
Regarding claim 12, AHMED teaches the processor is configured to access the document embedding vector from the local memory (section 0031; query processor 116 generates an embedding vector for a received query. As noted above, data store 114 may store embedding vectors for documents, such that query processor 116 uses the trained model to process the query embedding vector in view of the document embedding vectors stored by data store 114, thereby generating a responsive set of documents).
Regarding claims 13 and 18, AHMED teaches a method, comprising:
identifying a query embedding vector at a processor (section 0023; a user of client device 104 may use client application 120 to identify a set of documents that are responsive to a search query. Client application 120 may receive a search query from a user, which may be provided to server device 102. Server device 102 may process the search query (e.g., using query processor 116) to determine a set of documents that is responsive to the search query.);
locating a document embedding vector in a local memory of the processor, a memory, or a cache-coherent interconnect storage device using an Embedding Management Unit (EMU) (section 0024; documents identified as being responsive to a search query are stored or otherwise provided by a data source, such as data source 104. For example, document store 118 of data source 104 may store any of a variety of documents, including, but not limited to, text documents, audio files, video files, and/or webpages of a website);
processing the query embedding vector and the document embedding vector to produce a result (section 0031; query processor 116 generates an embedding vector for a received query. As noted above, data store 114 may store embedding vectors for documents, such that query processor 116 uses the trained model to process the query embedding vector in view of the document embedding vectors stored by data store 114, thereby generating a responsive set of documents); and
sending a document based at least in part on the result (section 0031; a dot-product ANN search may be used, such that query processor 116 generates a set of documents responsive to the query that may be returned to client device 106. The returned set of documents may comprise references to the identified documents and/or excerpts from the documents).
AHMED does not clearly teach a query is received from a host and send a document to the host.
However, Sims teaches a query is received from a host and send a document to the host (col.20, lines 12-18; The document identification process 600 may be invoked in various contexts. The invocation context will determine where the input query is received from and where the response is returned. Typically, however, the input query will be received from either a user device 110 (as configured by a client application 112) or a collaboration system 120 (as configured by a server application 122)).
It would have been obvious to the ordinary skill in the art before the effective filing date of the claimed invention to utilize the teachings Sims into AHMED such as a query is received from a host and send a document to the host because this will identify documents that relate to an input query (col.2, lines 23-24) and facilitate the storage and retrieval of data (col.4, lines 41-43).
AHMED and Sims do not clearly teach the EMU tracks where the document embedding vector is stored in the local memory, the memory, or the cache-coherent interconnect storage device.
However, Jacobs teaches the EMU tracks where the document embedding vector is stored in the local memory, the memory, or the cache-coherent interconnect storage device (section 0013; 0035; 0043 and section 0073; the document embedding vectors computed by the document embedding component can be cached and be re-used; The document embedding component 140 can be configured to compute, for each document, an n-dimensional vector of real numbers. The n-dimensional vector can be herein referred to as the document embedding vector; the computed n-dimensional vectors of the documents can also be cached).
It would have been obvious to the ordinary skill in the art before the effective filing date of the claimed invention to utilize the teachings Jacobs into AHMED and Sims such as the EMU tracks where the document embedding vector is stored in the local memory, the memory, or the cache-coherent interconnect storage device because it efficiently compute approximate embedding vectors of newly added documents. Thus, new documents can be accommodated in the system without the need to immediately re-train the system (section 0013 of Jacobs).
Regarding claims 14 and 19, AHMED teaches locating the document embedding vector in the local memory of the processor, the memory, or the cache-coherent interconnect storage device using the EMU includes locating the document embedding vector in the memory or the cache-coherent interconnect storage device using EMU (Fig.5; it is taught as the operating system 505, for example, may be suitable for controlling the operation of the computing device 500; section 0031; query processor 116 generates an embedding vector for a received query. As noted above, data store 114 may store embedding vectors for documents, such that query processor 116 uses the trained model to process the query embedding vector in view of the document embedding vectors stored by data store 114, thereby generating a responsive set of documents).
Regarding claim 15, AHMED teaches processing the query embedding vector and the document embedding vector to produce the result includes processing the query embedding vector and the document embedding vector in the memory or the cache-coherent interconnect storage device to produce the result (section 0031; query processor 116 generates an embedding vector for a received query. As noted above, data store 114 may store embedding vectors for documents, such that query processor 116 uses the trained model to process the query embedding vector in view of the document embedding vectors stored by data store 114, thereby generating a responsive set of documents).
Regarding claims 16 and 20, AHMED teaches processing the query embedding vector and the document embedding vector to produce the result includes: copying the document embedding vector from the memory or the cache-coherent interconnect storage device into the local memory(Accordingly, query processor 116 uses the trained model to process queries (e.g., as may be received from client application 120 of client device 106) and generate a set of documents that is responsive to the query accordingly. As an example, query processor 116 generates an embedding vector for a received query. As noted above, data store 114 may store embedding vectors for documents, such that query processor 116 uses the trained model to process the query embedding vector in view of the document embedding vectors stored by data store 114, thereby generating a responsive set of documents. For example, a dot-product ANN search may be used, such that query processor 116 generates a set of documents responsive to the query that may be returned to client device 106); and processing the query embedding vector and the document embedding vector in the local memory to produce the result (section 0031; a dot-product ANN search may be used, such that query processor 116 generates a set of documents responsive to the query that may be returned to client device 106. The returned set of documents may comprise references to the identified documents and/or excerpts from the documents).
Claim 2 is rejected under 35 U.S.C. 103 as being unpatentable over AHMED et al., US 2022/0179871 A1, Sims., US 10,565234 B1 and Jacobs et al., US 2023/0244990 A1, further in view of JUNG et al., US 2024/0012684 A1.
Regarding claim 2, AHMED, Sims and Jacobs teach the claimed invention as shown above, AHMED, Sims and Jacobs do not clearly teach the cache-coherent interconnect storage device includes a Compute Express Link (CXL) storage device. However, JUNG teaches the cache-coherent interconnect storage device includes a Compute Express Link (CXL) storage device (claim 1 and section 0118-0119; a host server and a memory device connected through a compute express link (CXL) network, wherein a computing complex of the host server is connected to a memory resource of the memory device through a CXL packet transmitted through the CXL network). It would have been obvious to the ordinary skill in the art before the effective filing date of the claimed invention to utilize the teachings JUNG into AHMED, Sims and jacobs such as the cache-coherent interconnect storage device includes a Compute Express Link (CXL) storage device because if the host server 100D is connected to the HDM through the CXL switch 300D having the computing ability, only a result computed by the CXL switch 300D may be used without a need for accessing all scattered embedding vectors (section 0119 of JUNG).
Claim 3 is rejected under 35 U.S.C. 103 as being unpatentable over AHMED et al., US 2022/0179871 A1, Sims., US 10,565234 B1 and Jacobs et al., US 2023/0244990 A1, further in view of Kalyanasundharam et al., US 2023/0195662 A1.
Regarding claim 3, AHMED, Sims and Jacobs teach the claimed invention as shown above, AHMED, Sims and Jacobs do not clearly teach the local memory, the memory, and the cache-coherent interconnect storage device form a unified memory. However, Kalyanasundharam teaches the local memory, the memory, and the cache-coherent interconnect storage device form a unified memory (section 0023; data processing system 300 implements a unified memory space in which all memory in the system is potentially visible to each memory accessing agent such as CPU core complexes 311).
It would have been obvious to the ordinary skill in the art before the effective filing date of the claimed invention to utilize the teachings Kalyanasundharam into AHMED, Sims and Jacobs such as the local memory, the memory, and the cache-coherent interconnect storage device form a unified memory because data processing system implements a unified memory space will provide a highly integrated, high-performance digital data processor that performs many of the functions associated with a workstation, a server, or the like (section 0023 of Kalyanasundharam).
Claim 5-7 are rejected under 35 U.S.C. 103 as being unpatentable over AHMED et al., US 2022/0179871 A1, Sims., US 10,565234 B1 and Jacobs et al., US 2023/0244990 A1, further in view of Liu et al., US 2021/0081389 A1.
Regarding claims 5-7, AHMED, Sims and Jacobs teach the claimed invention as shown above, AHMED, Sims and Jacobs do not clearly teach the EMU is further configured to evict a second document embedding vector from the local memory using an eviction policy; the EMU is configured to copy the second document embedding vector from the local memory to the memory using the eviction policy; the EMU is further configured to evict a second document embedding vector from the memory using an eviction policy. However, Liu teaches the EMU is further configured to evict a second document embedding vector from the local memory using an eviction policy; the EMU is configured to copy the second document embedding vector from the local memory to the memory using the eviction policy; the EMU is further configured to evict a second document embedding vector from the memory using an eviction policy (section 0083; section 0093 and section 0145; in addition to residing on disk, database blocks 372-375, table rows 312-319, and JSON documents 352-359 reside in memory 304 as shown. Within memory 304 and as shown, database blocks 372-375 reside in buffer cache 302 that can transfer database block(s) to/from disk. Thus, table rows 312-321, and JSON documents 352-359 reside in buffer cache 302 as shown. In this example, buffer cache 302 has an eviction policy such as least recently used (LRU)). It would have been obvious to the ordinary skill in the art before the effective filing date of the claimed invention to utilize the teachings Liu into AHMED, Sims and Jacobs such as the EMU is further configured to evict a second document embedding vector from the local memory using an eviction policy; the EMU is configured to copy the second document embedding vector from the local memory to the memory using the eviction policy; the EMU is further configured to evict a second document embedding vector from the memory using an eviction policy because the identifier vector and the JSON document vector may have parallel offsets, such that if a JSON document's offset into the JSON document vector is known, then that offset can be used in the identifier vector to find the identifier of the table row that stores that JSON document (section 0093 of Liu).
Claim 21 is rejected under 35 U.S.C. 103 as being unpatentable over AHMED et al., US 2022/0179871 A1, Sims., US 10,565234 B1 and Jacobs et al., US 2023/0244990 A1., further in view of Laycock et al., US 2012/0079211 A1.
Regarding claim 21, AHMED, Sims and Jacobs teach the claimed invention as shown above, AHMED, Sims and Jacobs do not clearly teach the cache-coherent interconnect storage device maintains coherency between the cache-coherent storage device, the local memory and the memory. However, Laycock teaches the cache-coherent interconnect storage device maintains coherency between the cache-coherent storage device, the local memory and the memory (section 0009 and section 0034; a cache for storing a local copy of at least one of said data items stored in said memory, said at least one initiator being configured to maintain coherency of said locally stored at least one data item, by in response to said locally stored data item being updated, marking said locally stored data item as dirty until said at least one initiator has performed a writeback operation and written said updated data item to said memory). It would have been obvious to the ordinary skill in the art before the effective filing date of the claimed invention to utilize the teachings Laycock into AHMED, Sims and Jacobs such as the cache-coherent interconnect storage device maintains coherency between the cache-coherent storage device, the local memory and the memory because it provides an effective power efficient and fast way of ensuring coherency of writebacks (section 0035 of Laycock).
When responding to the office action, Applicant is advised to clearly point out the patentable novelty which he or she thinks the claims present in view of the state of the art disclosed by the references cited or the objections made. He or she must also show how the amendments avoid such references or objections. See 37 C.F.R. 1.111 (c).
When responding to the office action, Applicants are advised to provide the examiner with the line numbers and page numbers in the application and/or references cited to assist examiner to locate the appropriate paragraphs.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to HUA JASMINE SONG whose telephone number is (571)272-4213. The examiner can normally be reached on 9:00am to 5:30pm.
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) at http:/Wwww.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, ROCIO DEL MAR PEREZ-VELEZ can be reached on 571-270-5935. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer
Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/HUA J SONG/Primary Examiner, Art Unit 2133