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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on January 7th 2026 has been entered.
Response to Amendment
3. The Amendment filed on January 7th 2026 has been entered. Claims 1 and 11 have been amended and claims 1 - 11 are pending in the application.
Response to Arguments
35 U.S.C. §101
4. Applicant's arguments, see Remarks pp. 6 - 13, filed January 7th 2026, with
respect to the rejections of claims 1 - 11 under 35 U.S.C. §101 have been fully
considered but they are not persuasive.
Applicant argues that the claims in the present invention are not directed to a mental process as indicated for at least three reasons, 1) the claims are necessarily rooted in computer technology, 2) the claimed invention is directed to an improvement and 3) the claims amount to significantly more than the judicial exception.
First, applicant argues by citing DDR Holdings v Hotels.com , L.P. Case No. 2013-1505, at 20 (Fed. Cir. 2014) and compares and absorbs its recitation, “the claimed solution is necessarily rooted in compute technology in order to overcome a problem specifically arising in the realm of computer networks” emphasis applicants. Applicant argues further that instead of executing instructions in sequence, an intermediate representation is made to transform and execute the same with the solution being a way to optimize programs without creating an iterative structure. Thus, the claims are directed to a concrete and non-theoretical concept that differs significantly from the examples laid out in Alice and in the HirshfiedMemo.
Examiner respectfully agrees in part and disagree. The claimed invention may be rooted in computer technology but that not sine qua non in the disposition of whether the claimed invention is patent eligible. Indeed the Courts have held, claimed inventions being patent ineligible when a generic computer is used to engage in functions that can be employed by the human mind. "It is well-settled that mere recitation of concrete, tangible components is insufficient to confer patent eligibility to an otherwise abstract idea"). The programmed computer or "special purpose computer" test of In re Alappat, 33 F.3d 1526, 31 USPQ2d 1545 (Fed. Cir. 1994) (i.e., the rationale that an otherwise ineligible algorithm or software could be made patent-eligible by merely adding a generic computer to the claim for the "special purpose" of executing the algorithm or software) was also superseded by the Supreme Court’s Bilski and Alice Corp. decisions. Eon Corp. IP Holdings LLC v. AT&T Mobility LLC, 785 F.3d 616, 623, 114 USPQ2d 1711, 1715 (Fed. Cir. 2015) ("[W]e note that Alappat has been superseded by Bilski, 561 U.S. at 605–06, and Alice Corp. v. CLS Bank Int’l, 573 U.S. 208, 110 USPQ2d 1976 (2014)"); Intellectual Ventures I LLC v. Capital One Bank (USA), N.A., 792 F.3d 1363, 1366, 115 USPQ2d 1636, 1639 (Fed. Cir. 2015) ("An abstract idea does not become nonabstract by limiting the invention to a particular field of use or technological environment, such as the Internet [or] a computer"). Thus it is not patent eligible.
Second, the applicant argues that the claimed invention is directed to an improvement, if the additional limitations reflect an improvement in the function of a computer, or an improvement to another technology or technical field, the claim integrates the judicial exception into a practical application and thus imposes and meaningful limit on the judicial exception. No further analysis is required. Emphasis the applicant’s.
Examiner agrees in part but disagrees in part. An improvement may be determined on the additional limitations reflect an improvement in the function of a computer, or an improvement to another technology or technical field, however, the examiner’s position on the claimed subject matter is it does not impose meaningfully limits the claim by going beyond generally linking the use of the judicial exception to a particular technological environment, and thus transforms a claim into patent-eligible subject matter. Within an intermediate representation, values are interchange or updated or modified and when the nodes converge a rule indicates its being “optimized” and stops further iterations. A threshold point reached based on condition is mere form of data gathering. Thus it is not patent eligible.
Third, applicant argues that the ordering combination amounts to significantly more.
Examiner respectfully disagrees. All the claim elements were examined individually and in combination, such that the claim as a whole did not amount to more than the judicial exception. The additional claims were not significantly more. As held by the Courts, “generally linking the use of the judicial exception to a particular technological environment or field of use, e.g., a claim describing how the abstract idea of hedging could be used in the commodities and energy markets, as discussed in Bilski v. Kappos, 561 U.S. 593, 595, 95 USPQ2d 1001, 1010 (2010).
35 U.S.C. §103
5. Applicant's arguments, see Remarks pp. 13 -16, filed January 7th 2026, with
respect to the rejections of claims 1-20 under 35 U.S.C. §103 have been fully
considered and they are persuasive.
Applicant argues that the claimed features of independent claim 1 together wits amendments are not taught by the combination of recited arts.
Examiner respectfully agrees and submits newly found art in relation to the claimed features and subsequent amendments, therefore upon further consideration new grounds of rejection have been necessitated due to Applicant's amendments and are made in view of Li et al., (United States Patent Publication Number 20230033019) hereinafter Li and Boettcher et al., (United States Patent Publication Number 2018/0315228) hereinafter Boettcher
Claim Rejections – 35 U.S.C. §101
6. 35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1 – 11 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e., a law of nature , a natural phenomenon or an abstract idea.) without significantly more.
The claims are analyzed for subject matter eligibility using a two-part subject matter eligibility analysis (MPEP 2016).
Claim 1 is rejected under 35 U.S.C. 101 because the claimed invention is
directed to an abstract idea without significantly more. The claim recites, “determining the presence or absence of an in-place operation based on the extracted information on data and the extracted information on operation;” These limitations may be properly identified as reciting the abstract idea of a “mental processes.” Mental processes – concepts performed in the human mind (including an observation, evaluation, judgment or opinion) falls within the grouping of abstract ideas, see MPEP 2106.04(a)(2).
The courts consider a mental process (thinking) that “can be performed in the human mind, or by a human using a pen and paper” to be an abstract idea. CyberSource Corp. v. Retail Decisions, Inc., 654 F.3d 1366, 1372, 99 USPQ2d 1690, 1695 (Fed. Cir. 2011). As the Federal Circuit explained, "methods which can be performed mentally, or which are the equivalent of human mental work, are unpatentable abstract ideas the ‘basic tools of scientific and technological work’ that are open to all.’" 654 F.3d at 1371, 99 USPQ2d at 1694 (citing Gottschalk v. Benson, 409 U.S. 63, 175 USPQ 673 (1972)). See also Mayo Collaborative Servs. v. Prometheus Labs. Inc., 566 U.S. 66, 71, 101 USPQ2d 1961, 1965 ("‘[M]ental processes[] and abstract intellectual concepts are not patentable, as they are the basic tools of scientific and technological work’" (quoting Benson, 409 U.S. at 67, 175 USPQ at 675)); Parker v. Flook, 437 U.S. 584, 589, 198 USPQ 193, 197 (1978) (same).
This judicial exception is not integrated into a practical application because the additional recitations of, “wherein the in-place operation is an operation in which a value stored in a memory allocated to input data is replaced with a value of output data; and if there is the in-place operation, creating an intermediate representation including nodes representing the data and the operation and edges representing an input and output relationship between the data and the operation using the extracted information on data, the extracted information on operation, and a creation rule associated with the in-place operation, wherein the creation rule is for creating and updating a new output data node in the intermediate representation such that the input data of the in-place operation is data that is replaced with new output data node after the in-place operation, and wherein the new output data node corresponds to the information on data for the output and the creation rule prevents an iterative structure from forming” are insignificant pre-processing step of replacing data in memory recited at a high level of generality, creating nodes and edges are mental processes that can be achieved using a paper and pen, observing the implementation of a rule that replaces data can be achieved mentally using a paper. These steps are mere data gathering steps.
Thus the limitations when considered individually and in combination, i.e. observation of an operation in an information element coupled with the evaluation of the presence of the operation to create a representation with a rule are mental processes that can be reduced by a pen and a paper. These additional elements are not sufficient to amount to significantly more than the judicial exception since they amount to simply implementing the abstract idea on a database and thus not integrated into a practical solution.
The claim recites extraction of information from a program via a processor. This is recited at a high level of generality that when considered individually and combination with the observation of an operation in an information element coupled with the evaluation of the presence of the operation to create a representation with a rule are well-known, routine and conventional. Obtaining data elements from a database are well-understood, routine, conventional computer functions as recognized by the court decisions such as storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363, 115 USPQ2d at 1092-93. Reference is made to Shai et al., (United States Patent Publication Nubmer 20190354544) which teaches in paragraph [0080] “Each node 60, 62, 64, 66, 68, 70 and 72 of the graph represents an entity identified from one or more of the set of documents, and vertices (e.g., edges) of each node represent an association (e.g., relationship) between entities” and in paragraph [0336] “A PermID is allocated; Published. All the defining characteristics are confirmed, a PermID has been allocated, and the content can be sent out in a strategic data interface; Deleted. This state can only be applied if the content has not been published; Superseded. Replaced by another one”
These claim limitations, under their broadest reasonable interpretation, covers mental processes. If a claim limitation, under its broadest reasonable interpretation, covers mental processes , then it falls within the “Mental Process” grouping of abstract ideas. Accordingly, the claims recites an abstract idea. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. Thus, the claim is not patent eligible.
Claim 2 is rejected under 35 U.S.C. 101 because the claimed invention is
directed to an abstract idea without significantly more. The claim depends from
independent claim 1 and does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim recites, “storing, in a database, the extracted data for input and output, and a corresponding relationship between data included in the intermediate representation.” This additional element of storing data in a database is a mere step of data gathering and is not sufficient to amount to significantly more than the judicial exception since they amount to simply
implementing the abstract idea on a database and thus not integrated into a practical solution.
The additional element storing data in a database when considered singularly and in combination with the aforementioned elements are well-understood, routine, conventional computer functions as recognized by the court decisions such as storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363, 115 USPQ2d at 1092-93.
These claim limitations, under their broadest reasonable interpretation, covers mental processes. If a claim limitation, under its broadest reasonable interpretation, covers mental processes , then it falls within the “Mental Process” grouping of abstract ideas. Accordingly, the claims recites an abstract idea. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. Thus, the claim is not patent eligible.
Claim 3 is rejected under 35 U.S.C. 101 because the claimed invention is
directed to an abstract idea without significantly more. The claim depends from
dependent claim 2 and does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim recites, “wherein the creating the intermediate representation includes creating the new intermediate representation by creating output data of an operation of the intermediate representation corresponding to output data of the in-place operation, and a data name of the output data of the operation of the intermediate representation is different from a data name of the input data of the in-place operation.” Observing the differences of the names of data is a mental process and creating data for output is a mere data gathering step.
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception when considered singularly or in combination because observing the differences types of data by their “headings” as input or output is a form of sorting or distinguishing the data and creating data for output are insignificant extra solution activities. The courts has held that, “Arranging a hierarchy of groups, sorting information, eliminating less restrictive pricing information and determining the price, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1331, 115 USPQ2d 1681, 1699 (Fed. Cir. 2015)” are well-understood, routine and conventional activity.
These claim limitations, under their broadest reasonable interpretation, covers mental processes. If a claim limitation, under its broadest reasonable interpretation, covers mental processes , then it falls within the “Mental Process” grouping of abstract ideas. Accordingly, the claims recites an abstract idea. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. Thus, the claim is not patent eligible.
Claim 4 is rejected under 35 U.S.C. 101 because the claimed invention is
directed to an abstract idea without significantly more. The claim depends from
dependent claim 3 and does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim recites, “wherein the storing, in the database, includes updating the database by changing such that a pointer of the input data of the in-place operation points to the new output data of the operation of the intermediate representation corresponding to the output data of the in-place operation.” Storing data in a database is a step of mere data gathering and does not impose a meaningful limit on the abstract idea. This step when considered individually and in combination with the aforementioned steps does not integrate the judicial exception into a practical solution. Storing data in a database are well-understood, routine, conventional computer functions as recognized by the court decisions such as storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363, 115 USPQ2d at 1092-93.
These claim limitations, under their broadest reasonable interpretation, covers mental processes. If a claim limitation, under its broadest reasonable interpretation, covers mental processes , then it falls within the “Mental Process” grouping of abstract ideas. Accordingly, the claims recites an abstract idea. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. Thus, the claim is not patent eligible.
Claim 5 is rejected under 35 U.S.C. 101 because the claimed invention is
directed to an abstract idea without significantly more. The claim depends from
dependent claim 2 and does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim recites, “wherein the in-place operation includes a first in-place operation and a second in-place operation following the first in-place operation, the creating the intermediate representation includes creating the intermediate representation by creating the output data of the operation of the intermediate representation corresponding to the output data of the first in-place operation and the output data of the second in-place operation, and a data name of the input data of the first in-place operation, a data name of the output data of the first in-place operation, and a data name of the output data of the second in-place operation are different from one another.” Observing the composition of operations and the names of data inputs and types are mental processes and the arrangement data operations in terms of their scheduled interrelationships are a mere form of data gathering that does not impose a meaningful limit on the abstract idea. Therefore the claimed limitations when considered singularly and in combination are not integrated into a practical solution. The courts has held that, “Arranging a hierarchy of groups, sorting information, eliminating less restrictive pricing information and determining the price, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1331, 115 USPQ2d 1681, 1699 (Fed. Cir. 2015)” are well-understood, routine and conventional activity.
These claim limitations, under their broadest reasonable interpretation, covers mental processes. If a claim limitation, under its broadest reasonable interpretation, covers mental processes , then it falls within the “Mental Process” grouping of abstract ideas. Accordingly, the claims recites an abstract idea. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. Thus, the claim is not patent eligible.
Claim 6 is rejected under 35 U.S.C. 101 because the claimed invention is
directed to an abstract idea without significantly more. The claim depends from
dependent claim 5 and does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim recites, “wherein the storing, in the database, includes updating the database by changing such that a pointer of the input data of the first in-place operation points to the output data of the operation of the intermediate representation corresponding to the output data of the second in-place operation.” Storing data in a generic database by pointer location is recited at a high level of generality and is therefore mere data gathering, thus does not impose a meaningful limit on the abstract idea. This step of storing data when considered singularly or in combination does not integrate the abstract idea into a practical solution. Obtaining data elements from a database are well-understood, routine, conventional computer functions as recognized by the court decisions such as storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363, 115 USPQ2d at 1092-93
These claim limitations, under their broadest reasonable interpretation, covers mental processes. If a claim limitation, under its broadest reasonable interpretation, covers mental processes , then it falls within the “Mental Process” grouping of abstract ideas. Accordingly, the claims recites an abstract idea. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. Thus, the claim is not patent eligible.
Claim 7 is rejected under 35 U.S.C. 101 because the claimed invention is
directed to an abstract idea without significantly more. The claim depends from
dependent claim 2 and does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim recites, “wherein the storing, in the database, includes: if a first size of the input data of the in-place operation is different from a second size of the output data of the in-place operation, updating the database by changing such that a pointer of the input data of the in-place operation points to the output data of the operation of the intermediate representation corresponding to the output data of the in-place operation; and storing information on an operation to return to the first size, in association with the input data of the in-place operation.” Observing or evaluating the size of data elements is a cognitive process that can be reduced by paper and pen. Further storing data in a generic database recited at a high level of generality is a mere step of data gathering. The claimed limitations when considered singularly and in combination do not impose a meaningful limit on the abstract idea and thus do not integrate it into a practical solution. Obtaining data elements from a database are well-understood, routine, conventional computer functions as recognized by the court decisions such as storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363, 115 USPQ2d at 1092-93
These claim limitations, under their broadest reasonable interpretation, covers mental processes. If a claim limitation, under its broadest reasonable interpretation, covers mental processes , then it falls within the “Mental Process” grouping of abstract ideas. Accordingly, the claims recites an abstract idea. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. Thus, the claim is not patent eligible.
Claim 8 is rejected under 35 U.S.C. 101 because the claimed invention is
directed to an abstract idea without significantly more. The claim depends from
dependent claim 7 and does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim recites, “creating the intermediate representation includes, as a subsequent operation of the in-place operation, if there is an operation using the first size of the input data of the in-place operation, creating the a second intermediate representation using the operation to return the second size of the input data of the in-place operation to the first size.” Comparing the size of data elements and then repositioning a process flow to return a desired size of data is a cognitive process that results in the mere data gathering of data elements. This is an insignificant extra solution activity that does not impose a meaningful limit on the abstract idea. This claimed step when considered singularly and in combination does not integrate the abstract idea into a practical solution.
These claim limitations, under their broadest reasonable interpretation, covers mental processes. If a claim limitation, under its broadest reasonable interpretation, covers mental processes , then it falls within the “Mental Process” grouping of abstract ideas. Accordingly, the claims recites an abstract idea. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. Thus, the claim is not patent eligible.
Claim 9 is rejected under 35 U.S.C. 101 because the claimed invention is
directed to an abstract idea without significantly more. The claim depends from
dependent claim 8 and does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim recites, “wherein the storing, in the database, includes: if the subsequent operation of the in-place operation is a subsequent in-place operation, changing the second size of the output data of the in-place operation to the first size by pointing each pointer of the input data of the in-place operation and the output data of the in-place operation to the output data of the operation of the second intermediate representation corresponding to the output data of the subsequent in-place operation; and storing information to return to the second size, in association with the output data of the in-place operation.” Storing information into a generic database is recited at a high level of generality that it does not impose a meaningful limit on the abstract idea. The step when considered singularly and in combination does not impose a meaningful limit on the abstract idea. Obtaining data elements from a database are well-understood, routine, conventional computer functions as recognized by the court decisions such as storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363, 115 USPQ2d at 1092-93
These claim limitations, under their broadest reasonable interpretation, covers mental processes. If a claim limitation, under its broadest reasonable interpretation, covers mental processes , then it falls within the “Mental Process” grouping of abstract ideas. Accordingly, the claims recites an abstract idea. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. Thus, the claim is not patent eligible.
Independent claim 10 corresponds to independent claim 1 but for the recitation of, “A non-transitory computer-readable recording medium storing instructions that, when executed by one or more processors.” These claim limitations, under their broadest reasonable interpretation, covers mental processes but for the recitation of a non-transitory computer-readable recording medium storing instructions that, when executed by one or more processors. That is, other than reciting "an apparatus ..," nothing in the claim element precludes the step from practically being performed in the mind. If a claim limitation, under its broadest reasonable interpretation, covers mental processes , then it falls within the “Mental Process” grouping of abstract ideas. Accordingly, the claims recites an abstract idea. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. Thus, the claim is not patent eligible.
Independent claim 11 corresponds to independent claim 1 but for the recitation of, “system, comprising: a memory; and one or more processors connected to the memory and configured to execute one or more computer-readable programs included in the memory.” These claim limitations, under their broadest reasonable interpretation, covers mental processes but for the recitation of system, comprising: a memory; and one or more processors connected to the memory and configured to execute one or more computer-readable programs included in the memory. That is, other than reciting " system, comprising: a memory; and one or more processors connected to the memory and configured to execute one or more computer-readable programs included in the memory ..," nothing in the claim element precludes the step from practically being performed in the mind. If a claim limitation, under its broadest reasonable interpretation, covers mental processes , then it falls within the “Mental Process” grouping of abstract ideas. Accordingly, the claims recites an abstract idea. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. Thus, the claim is not patent eligible.
Claim Rejections – 35 U.S.C. §103
7. 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.
8. The factual inquiries set forth in Graham v John Deere Co., 383 U.S. 1, 148 USPQ
459 (1966), that are applied for establishing a background for determining obviousness
under 35 U.S.C. 103 are summarized as follows:
a. Determining the scope and contents of the prior art
b. Ascertaining the differences between the prior art and the claims at issue
c. Resolving the level of ordinary skill in the pertinent art
d. Considering objective evidence present in the application indicating
obviousness or nonobviousness
Claims 1, 10 and 11 are rejected under 35 U.S.C. 103 as being unpatentable over Boettcher et al., (United States Patent Publication Number 2018/0315228) hereinafter Boettcher in view of Li et al., (United States Patent Publication Number 20230033019) hereinafter Li
Regarding claim 1 Boettcher teaches a method for creating an intermediate representation (//update depending on edge strategy
case picking:
pickedEdge-NILEDGE
for every edge each in oldGraph
if isMultiEdge(
edge.destination,lastDestination)
pickedEdge-pickingFunction.pick(
else
pickedEdge,edge [0116]) such as “intermediate representation” from a program (pseudo-coded harness [0116]) including an in-place operation, ("inPlace": true, II we decided to do the mutation inplace [0123]) wherein the in-place operation ("inPlace": true, II we decided to do the mutation inplace [0123])
Boettcher does not fully disclose wherein the in-place operation is an operation in which a value stored in a memory allocated to input data is replaced with a value of output data, the method being performed by one or more processors and comprising: extracting, from the program, information on data for input and output and information on operation; determining the presence of an in-place operation based on the extracted information on data and the extracted information on operation; and if there is the in-place operation, creating an intermediate representation including nodes representing the data and the operation and edges representing an input and output relationship between the data and the operation using the extracted information on data, the extracted information on operation, and a creation rule associated with the in-place operation, wherein the creation rule is for creating and updating a new output data node in the intermediate representation such that the input data of the in-place operation is data that is replaced with the new output data node after the in-place operation, and wherein the new output data node corresponds to the information on data for the output and the creation rule prevents an iterative structure from forming.
Li teaches is an operation in which a value stored in a memory allocated to input data is replaced with a value of output data, (ReadMessage is replaced with WriteMessage, and WriteMessage is simultaneously reset to
prepare for a next round of PS read-write [0130]) the method (Figs. 2 and 3 methods [0014], [0015]) being performed by one or more processors (processor apparatuses [0026]) and comprising: extracting, from the program, information on data for input and output (A core degree ( core value) of the node may be input to a downstream machine learning task as a topological feature to implement a activity model mining task to recognize whether the node in the graph network model is a user or a follower [0034]) and information on operation; (data mining may be performed based on the graph network model to detect whether there is an unusual interaction behavior [0034]) SEE ALSO [0055] determining the presence of an in-place operation (ABS., iteratively updating node core degrees of all or some of the nodes in the relationship graph network) (Fig. 2, (S220) iteratively update node core degrees of all or part of the nodes in the relationship graph network [0048]) such as “in-place operation” based on the extracted information on data and the extracted information on operation; (data generated in social contacts between multiple interacting objects is collected, and then the multiple interacting objects and interactive relationships between these interacting objects may be extracted [0055]) and if there is the in-place operation, (ABS., iteratively updating node core degrees of all or some of the nodes in the relationship graph network) (Fig. 2, (S220) iteratively update node core degrees of all or part of the nodes in the relationship graph network [0048]) such as “in-place operation” creating an intermediate representation including nodes representing the data and the operation and edges representing an input and output relationship between the data (Fig. 8 After a first iteration round, core values of a part of nodes are updated, and in this part of nodes whose core values are updated, the minimum core value is minCoreC1) = 1. After a second iteration round, core values of another part of nodes are updated, and in this part of nodes whose core values are updated, the minimum core is minCoreC2) = 1. [0108]) such as “intermediate representation” and the operation using the extracted information on data, the extracted information on operation, (data generated in social contacts between multiple interacting objects is collected, and then the multiple interacting objects and interactive relationships between these interacting objects may be extracted [0055]) and a creation rule (According to a rule that a core value of a node decreases progressively in each iteration round, when minCore(t) > minCoreC1- 1), it indicates that all nodes whose core values are less than or equal to minCoreC1- 1l have converged at this time and may not be updated later. [0107]) such as “creation rule” associated with the in-place operation, (ABS., iteratively updating node core degrees of all or some of the nodes in the relationship graph network) (Fig. 2, (S220) iteratively update node core degrees of all or part of the nodes in the relationship graph network [0048]) such as “in-place operation” wherein the creation rule (According to a rule that a core value of a node decreases progressively in each iteration round, when minCore(t) > minCoreC1- 1), it indicates that all nodes whose core values are less than or equal to minCoreC1- 1l have converged at this time and may not be updated later. [0107]) such as “creation rule” is for creating and updating a new output data node in the intermediate representation such that the input data of the in-place operation (Fig. 8 After a first iteration round, core values of a part of nodes are updated, and in this part of nodes whose core values are updated, the minimum core value is minCoreC1) = 1. After a second iteration round, core values of another part of nodes are updated, and in this part of nodes whose core values are updated, the minimum core is minCoreC2) = 1. [0108]) such as “intermediate representation” is data that is replaced with the new output data node after the in-place operation, (Fig. 8 After a first iteration round, core values of a part of nodes are updated, and in this part of nodes whose core values are updated, the minimum core value is minCoreC1) = 1. After a second iteration round, core values of another part of nodes are updated, and in this part of nodes whose core values are updated, the minimum core is minCoreC2) = 1. [0108]) such as “intermediate representation” and wherein the new output data node corresponds to the information on data for the output and the creation rule (According to a rule that a core value of a node decreases progressively in each iteration round, when minCore(t) > minCoreC1- 1), it indicates that all nodes whose core values are less than or equal to minCoreC1- 1l have converged at this time and may not be updated later. [0107]) such as “creation rule” prevents (In ( 4), whether numMsgs is O is determined. When numMsgs is 0, it indicates that the core values of all the nodes are no longer updated, and iteration is stopped [0119]) an iterative structure (iterative updating of h-indexes, [0115]) such as “iterative structure” from forming (When a core value of a node is updated, an updated core value is less than the original core value. According to a rule that a core value of a node decreases progressively in each iteration round, when minCore(t) > minCoreC1- 1), it indicates that all nodes whose core values are less than or equal to minCoreC1- 1l have converged at this time and may not be updated later. [0107])
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Boettcher to incorporate the teachings of Li whereby wherein the in-place operation is an operation in which a value stored in a memory allocated to input data is replaced with a value of output data, the method being performed by one or more processors and comprising: extracting, from the program, information on data for input and output and information on operation; determining the presence of an in-place operation based on the extracted information on data and the extracted information on operation; and if there is the in-place operation, creating an intermediate representation including nodes representing the data and the operation and edges representing an input and output relationship between the data and the operation using the extracted information on data, the extracted information on operation, and a creation rule associated with the in-place operation, wherein the creation rule is for creating and updating a new output data node in the intermediate representation such that the input data of the in-place operation is data that is replaced with the new output data node after the in-place operation, and wherein the new output data node corresponds to the information on data for the output and the creation rule prevents an iterative structure from forming. By doing so the current node core degree may be replaced with the temporary node core degree. If they are the same, it indicates that the computing node needs not to be updated in the current iteration round. Li [0095]
Claims 10 and 11 correspond to claim 1 and are rejected accordingly
Claims 2 and 3 are rejected under 35 U.S.C. 103 as being unpatentable over Boettcher et al., (United States Patent Publication Number 2018/0315228) hereinafter Boettcher in view of Li et al., (United States Patent Publication Number 20230033019) hereinafter Li and in further view of Lee et al., (Korean Patent Publication Number 10-2490539) hereinafter lee
Regarding claim 2 Boettcher in view of Li teaches the method according to claim 1,
Boettcher as modified does not fully disclose storing, in a database, the extracted data for input and output, and a corresponding relationship between data included in the intermediate representation.
Lee teaches comprising storing, in a database, (The processor 100 may transmit the generated unit operation list to the intermediate expression DB 260 Page 9 ) such as “storing in a database” the extracted data for input and output,(extracted dependence of each unit operation Page 8) and a corresponding relationship between data (data dependent relationships Page 8)included in the intermediate representation (generated intermediate representation from the unitary operation list Page 8)
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Boettcher in view of Li to incorporate the teachings of Lee whereby storing, in a database, the extracted data for input and output, and a corresponding relationship between data included in the intermediate representation. By doing so the dataflow graph may include nodes representing scalar operations and edges representing data dependency relationships between scalar operations. Lee Pages 8 - 9
Regarding claim 3 Boettcher in view of Li and Lee teaches the method according to claim 2,
Boettcher as modified teaches of the in-place operation, (“inPlace” true [0123]) of the in-place operation. (“inPlace” true [0123])
Boettcher as modified does not fully disclose wherein the creating the intermediate representation includes creating the intermediate representation by creating the new output data of an operation of the intermediate representation corresponding to output data of the in-place operation, and a data name of the output data of the operation of the intermediate representation is different from a data name of the input data of the in-place operation.
Lee teaches wherein the creating the intermediate representation (The processor 100 may then generate an intermediate representation from the unitary operation list Page 8) includes creating the intermediate representation (The processor 100 may then generate an intermediate representation from the unitary operation list Page 8) by creating new output data (the result of A's operation is used as an input to B Page 5) such as A’s result is “output data” of an operation (unit operation A Page 5) of the intermediate representation (generated intermediate representation from the unitary operation list Page 8) corresponding to output data (the result of A's operation is used as an input to B Page 5) such as A’s result is “output data” and a data name of the output data (result of A's operation Page 5) of the operation (unit operation A Page 5) of the intermediate representation (generated intermediate representation from the unitary operation list Page 8) is different from a data name of the input data (input to B Page 5)
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Boettcher in view of Li to incorporate the teachings of Lee wherein the creating the intermediate representation includes creating the intermediate representation by creating the new output data of an operation of the intermediate representation corresponding to output data of the in-place operation, and a data name of the output data of the operation of the intermediate representation is different from a data name of the input data of the in-place operation. By doing so the processor 100 may generate an intermediate expression by concatenating one or more unit operation intermediate expressions, performing shared memory optimization, and distributing operations among threads. Lee Page 9.
Claim 4 is rejected under 35 U.S.C. 103 as being unpatentable over Boettcher et al., (United States Patent Publication Number 2018/0315228) hereinafter Boettcher in view of Li et al., (United States Patent Publication Number 20230033019) hereinafter Li, in view of Lee et al., (Korean Patent Publication Number 10-2490539) hereinafter lee and in further view of Jones et al., (United States Patent Publication Number 20210248115) hereinafter Jones
Regarding claim 4 Boettcher in view of Li and Lee teaches the method according to claim 2,
Boettcher as modified teaches of the in-place operation, (“inPlace” true [0123]) of the in-place operation. (“inPlace” true [0123])
Boettcher as modified does not fully disclose wherein the storing, in the database, includes updating the database by changing such that a pointer of the input data of the in-place operation points to the new output data of the operation of the intermediate representation corresponding to the output data of the in-place operation.
Lee as modified teaches, storing in the database(The processor 100 may transmit the generated unit operation list to the intermediate expression DB 260 Page 9 ) such as “storing in a database” includes input data (input to B Page 5) to the new output data (the result of A's operation is used as an input to B Page 5) such as A’s result is “output data” of the operation (unit operation A Page 5) of the intermediate representation corresponding to (generated intermediate representation from the unitary operation list Page 8) the output data (the result of A's operation is used as an input to B Page 5) such as A’s result is “output data”
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Boettcher in view of Li to incorporate the teachings of Lee wherein the storing, in the database, includes updating the database by changing such that a pointer of the input data of the in-place operation points to the new output data of the operation of the intermediate representation. By doing so the processor 100 may transmit the generated unit operation list to the intermediate expression DB 260 Lee Page 9
Jones teaches includes updating the database (Fig. 1, (104) updated optimized graph [0061]) such as “updating the dataset” by changing such that a pointer points (a parameter comprises a reference to a memory region or buffer [0054]) (an updated set of parameters 108 replace parameters 106 in an optimized graph 102. For example, parameters PA1 , PA2 , and PA3 106 used in optimization of an optimized graph 102 may be replaced by parameters PB1 , PB2 , and PB3 108 [0059], [0066], [0067]) SEE ALSO further parameter substitutions in paragraphs [0070] – [0073], [0075])
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Boettcher in view of Li and Lee to incorporate the teachings of Jones wherein includes updating the database by changing such that a pointer points to. By doing so some parameters are replaceable when efficacy of current optimizations is affected. Jones [0060].
Claim 5 is rejected under 35 U.S.C. 103 as being unpatentable over Boettcher et al., (United States Patent Publication Number 2018/0315228) hereinafter Boettcher in view of Li et al., (United States Patent Publication Number 20230033019) hereinafter Li and in further view of Lee et al., (Korean Patent Publication Number 10-2490539) hereinafter lee and in further view of Li et al., “The Deep Learning Compiler: A Comprehensive Survey” hereinafter Li-2
Regarding claim 5 Boettcher in view of Li and Lee teaches the method according to claim 2,
Boettcher as modified does not fully disclose wherein the in-place operation includes a first in-place operation and a second in-place operation following the first in-place operation, the creating the intermediate representation includes creating the intermediate representation by creating the output data of the operation of the intermediate representation corresponding to the output data of the first in-place operation and the output data of the second in-place operation, and a data name of the input data of the first in-place operation, a data name of the output data of the first in-place operation, and a data name of the output data of the second in-place operation are different from one another.
Lee teaches the creating the intermediate representation (The processor 100 may then generate an intermediate representation from the unitary operation list Page 8) includes creating the intermediate representation (The processor 100 may then generate an intermediate representation from the unitary operation list Page 8) by creating the output data (the result of A's operation is used as an input to B Page 5) such as A’s result is “output data” of the operation (unit operation A Page 5)of the intermediate representation (The processor 100 may then generate an intermediate representation from the unitary operation list Page 8)corresponding to the output data (the result of A's operation is used as an input to B Page 5) such as A’s result is “output data” and the output data(the result of A's operation is used as an input to B Page 5) such as A’s result is “output data” and a data name of the input data (input to B Page 5)and a data name of the output data (result of A's operation Page 5) are different from one another (input to B is different from result of A’s operation as titled operationally Page 5)
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Boettcher in view of Li to incorporate the teachings of Lee wherein the creating the intermediate representation includes creating the intermediate representation by creating the output data of the operation of the intermediate representation corresponding to the output data and the output data and a data name of the input data and a data name of the output data are different from one another. By doing so 'intermediate expression' may refer to a combination of one or more unit operation intermediate expressions generated based on data dependency relationships, and may be represented by, for example, DFS. Lee Page 8
Li-2 teaches wherein the in-place operation (placeholders for input and output such as tensor shape Page 14) such as “in-place operation” includes a first in-place operation and a second in-place operation following the first in-place operation, (the process is iterative Page 14) of the first in-place operation (placeholders for input and output such as tensor shape Page 14) such as “in-place operation” of the second in-place operation, (placeholders for input and output such as tensor shape Page 14) such as “in-place operation” (the process is iterative Page 14) and a data name of the input data of the first in-place operation, (placeholders for input and output such as tensor shape Page 14) such as “in-place operation” a data name of the output data (the result of the expression Page 11) of the first in-place operation, (placeholders for input and output such as tensor shape Page 14) such as “in-place operation” and a data name of the output data (the result of the expression Page 11) of the second in-place operation (placeholders for input and output such as tensor shape Page 14) such as “in-place operation” (the process is iterative Page 14)
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Boettcher in view of Li and Lee to incorporate the teachings of Li-2 wherein the in-place operation includes a first in-place operation and a second in-place operation following the first in-place operation, of the first in-place operation of the second in-place operation, and a data name of the input data of the first in-place operation,” a data name of the output data of the first in-place operation, and a data name of the output data of the second in-place operation. By doing so with the input data of a DL model, this kind of accelerators process the data through the different hardware blocks in the same sequence with layers. Additionally, with the streaming input data, all hardware blocks can be fully utilized in a pipeline manner. Li-2 Pages 6 – 7
Claims 6 - 9 are rejected under 35 U.S.C. 103 as being unpatentable over Boettcher et al., (United States Patent Publication Number 2018/0315228) hereinafter Boettcher in view of Li et al., (United States Patent Publication Number 20230033019) hereinafter Li, in view of Lee et al., (Korean Patent Publication Number 10-2490539) hereinafter lee, in view of Li et al., “The Deep Learning Compiler: A Comprehensive Survey” hereinafter Li-2 and in further view of Jones et al., (United States Patent Publication Number 20210248115) hereinafter Jones
Regarding clam 6 Boettcher in view of Li, Lee and Li-2 teaches the method according to claim 5,
Boettcher as modified does not fully disclose wherein the storing, in the database, includes updating the database by changing such that a pointer of the input data of the first in- place operation points to the output data of the operation of the intermediate representation corresponding to the output data of the second in-place operation.
Lee teaches wherein the storing, in the database(The processor 100 may transmit the generated unit operation list to the intermediate expression DB 260 Page 9 ) such as “storing in a database”
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Boettcher in view of Li to incorporate the teachings of Lee whereby storing, in a database,. By doing so the dataflow graph may include nodes representing scalar operations and edges representing data dependency relationships between scalar operations. Lee Pages 8 - 9
Li-2 teaches of the input data (as found in the Let expression that points to the operator and variable in the expression Page 11) of the first in-place operation (placeholders for input and output such as tensor shape Page 14) such as “in-place operation” to the output data (the result of the expression Page 11) of the operation of the intermediate representation (The intermediate representation (IR) is spread across both the frontend and the backend. Generally, IR is an abstraction of the program, and is used for program optimizations. Specifically, the DL models are translated into multi-level IRs in DL compilers, where the high-level IR resides in the frontend and the low-level IR resides in the backend. Page 9) SEE Fig 2 Page 10 corresponding to the output data (the result of the expression Page 11) of the second in-place operation (placeholders for input and output such as tensor shape Page 14) such as “in-place operation” (the process is iterative Page 14)
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Boettcher in view of Li and Lee to incorporate the teachings of Li-2 of the input data) of the first in-place operation to the output data of the operation of the intermediate representation corresponding to the output data of the second in-place operation. By doing so with the input data of a DL model, this kind of accelerators process the data through the different hardware blocks in the same sequence with layers. Additionally, with the streaming input data, all hardware blocks can be fully utilized in a pipeline manner. Li-2 Pages 6 - 7
Jones teaches includes updating the database (Fig. 1, (104) updated optimized graph [0061]) such as “updating the dataset” by changing such that a pointer points (a parameter comprises a reference to a memory region or buffer [0054]) (an updated set of parameters 108 replace parameters 106 in an optimized graph 102. For example, parameters PA1 , PA2 , and PA3 106 used in optimization of an optimized graph 102 may be replaced by parameters PB1 , PB2 , and PB3 108 [0059], [0066], [0067]) SEE ALSO further parameter substitutions in paragraphs [0070] – [0073], [0075])
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Boettcher in view of Li, Lee and Li-2 to incorporate the teachings of Jones wherein includes updating the database by changing such that a pointer points to. By doing so some parameters are replaceable when efficacy of current optimizations is affected. Jones [0060].
Regarding claim 7 Boettcher in view of Li and Lee teaches the method according to claim 2,
Boettcher as modified does not fully disclose wherein the storing, in the database, includes: if a first size of the input data of the in-place operation is different from a second size of the output data of the in-place operation, updating the database by changing such that a pointer of the input data of the in-place operation points to the output data of the operation of the intermediate representation corresponding to the output data of the in-place operation; and storing information on an operation to return to the first size, in association with the input data of the in-place operation.
Li-2 teaches if a first size (input shape Page 14) of the input data(as found in the Let expression that points to the operator and variable in the expression Page 11) of the in-place operation(placeholders for input and output such as tensor shape Page 14) such as “in-place operation” is different from a second size (output shape Page 14) of the output data (the result of the expression Page 11) of the in-place operation, (placeholders for input and output such as tensor shape Page 14) such as “in-place operation” of the input data (as found in the Let expression that points to the operator and variable in the expression Page 11) of the in-place operation (placeholders for input and output such as tensor shape Page 14) such as “in-place operation” the output data (the result of the expression Page 11) of the operation of the intermediate representation (The intermediate representation (IR) is spread across both the frontend and the backend. Generally, IR is an abstraction of the program, and is used for program optimizations. Specifically, the DL models are translated into multi-level IRs in DL compilers, where the high-level IR resides in the frontend and the low-level IR resides in the backend. Page 9) SEE Fig 2 Page 10 corresponding to the output data (the result of the expression Page 11) of the in-place operation; (placeholders for input and output such as tensor shape Page 14) such as “in-place operation” and storing information on an operation (TC stores the fastest known generated code version corresponded with given configuration by compilation cache and uses tuple as cache entry to present necessary information related to the version. Page 27) to return to the first size, (input shape Page 14) in association with the input data(as found in the Let expression that points to the operator and variable in the expression Page 11) of the in-place operation. (placeholders for input and output such as tensor shape Page 14) such as “in-place operation”
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Boettcher in view of Li to incorporate the teachings of Li-2 wherein if a first size of the input data of the in-place operation is different from a second size of the output data of the in-place operation, updating the database by changing such that a pointer of the input data of the in-place operation points to the output data of the operation of the intermediate representation corresponding to the output data of the in-place operation; and storing information on an operation to return to the first size, in association with the input data of the in-place operation. By doing so it is convenient for the programmer to change the input and output shape by using placeholders instead of changing the whole semantics. Li-2 Page 14
Jones teaches includes updating the database (Fig. 1, (104) updated optimized graph [0061]) such as “updating the dataset” by changing such that a pointer points (a parameter comprises a reference to a memory region or buffer [0054]) (an updated set of parameters 108 replace parameters 106 in an optimized graph 102. For example, parameters PA1 , PA2 , and PA3 106 used in optimization of an optimized graph 102 may be replaced by parameters PB1 , PB2 , and PB3 108 [0059], [0066], [0067]) SEE ALSO further parameter substitutions in paragraphs [0070] – [0073], [0075])
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Boettcher in view of Li, Lee and Li-2 to incorporate the teachings of Jones wherein includes updating the database by changing such that a pointer points to. By doing so some parameters are replaceable when efficacy of current optimizations is affected. Jones [0060].
Regarding claim 8 Boettcher in view of Li, Lee, Li-2 and Jones teaches the method according to claim 7,
Boettcher as modified does not fully disclose creating the intermediate representation includes, as a subsequent operation of the in-place operation, if there is an operation using the first size of the input data of the in-place operation, creating a second intermediate representation using the operation to return the second size of the input data of the in-place operation to the first size.
Lee as modified further teaches the creating the intermediate representation includes, (generated intermediate representation from the unitary operation list Page 8) creating the second intermediate representation (generated intermediate representation from the unitary operation list Page 8)
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Boettcher in view of Li, Li-2 and Jones to incorporate the teachings of Lee creating the intermediate representation includes, creating the second intermediate representation. By doing so an intermediate expression conversion module connects unit operation intermediate expressions. Lee page 7
Li-2 teaches subsequent operation of the in-place operation, (placeholders for input and output such as tensor shape Page 14) such as “in-place operation” (the process is iterative Page 14) if there is an operation (broadcasting Page 15) using the first size (input shape Page 14) of the input data (as found in the Let expression that points to the operator and variable in the expression Page 11) of the in-place operation, (placeholders for input and output such as tensor shape Page 14) such as “in-place operation” using the operation (broadcasting Page 15) to return the second size (output shape Pages 11, 14) of the input data (as found in the Let expression that points to the operator and variable in the expression Page 11) of the in-place operation (placeholders for input and output such as tensor shape Page 14) such as “in-place operation” to the first size. (input shape Page 14)
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Boettcher in view of Li, Lee, and Jones to incorporate the teachings of Li-2 wherein subsequent operation of the in-place operation, if there is an operation using the first size of the input data of the in-place operation, using the operation to return the second size of the input data of the in-place operation to the first size. By doing so with the input data of a DL model, this kind of accelerators process the data through the different hardware blocks in the same sequence with layers. Li-2 Page 6
Regarding claim 9 Boettcher in view of Li, Lee, Li-2 and Jones teaches the method according to claim 8,
Boettcher as modified does not fully disclose wherein the storing, in the database, includes: if the subsequent operation of the in-place operation is a subsequent in-place operation, changing the second size of the output data of the in-place operation to the first size by pointing each pointer of the input data of the in-place operation and the output data of the in-place operation to the output data of the operation of the second intermediate representation corresponding to the output data of the subsequent in-place operation; and storing information to return to the second size, in association with the output data of the in-place operation
Lee teaches wherein storing, in a database, (The processor 100 may transmit the generated unit operation list to the intermediate expression DB 260 Page 9 ) such as “storing in a database” of the operation of the second intermediate representation (generated intermediate representation from the unitary operation list Page 8)
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Boettcher in view of Li, Li-2 and Jones to incorporate the teachings of Lee wherein storing, in a database, of the operation of the second intermediate representation. By doing so an intermediate expression conversion module connects unit operation intermediate expressions. Lee page 7
Li-2 as modified teaches, includes: if the subsequent operation of the in-place operation is a subsequent in-place operation, (placeholders for input and output such as tensor shape Page 14) such as “in-place operation” (the process is iterative Page 14) changing (support for dynamic shapes Page 9) (TVM uses Any to represent an unknown dimension in order to support the dynamic model. The Any can be used in tensor type definition as one of the dimensions of the tensor shape Page 14) (Dynamic shape and control flow - Dynamic model becomes more and more popular in the field of deep learning, whose input shape or even model itself may change in execution Page 28) the second size (output shape Pages 11, 14) of the output data (the result of the expression Page 11) of the in-place operation(placeholders for input and output such as tensor shape Page 14) such as “in-place operation” to the first size(input shape Page 14) of the input data (as found in the Let expression that points to the operator and variable in the expression Page 11) of the in-place operation(placeholders for input and output such as tensor shape Page 14) such as “in-place operation” and the output data (the result of the expression Page 11) of the in-place operation (placeholders for input and output such as tensor shape Page 14) such as “in-place operation” to the output data (the result of the expression Page 11) corresponding to the output data(the result of the expression Page 11) of the subsequent in-place operation; (placeholders for input and output such as tensor shape Page 14) such as “in-place operation” (the process is iterative Page 14) and storing information (TC stores the fastest known generated code version corresponded with given configuration by compilation cache and uses tuple as cache entry to present necessary information related to the version. Page 27) to return to the second size, (output shape Pages 11, 14) in association with the output data (the result of the expression Page 11) of the in-place operation. (placeholders for input and output such as tensor shape Page 14) such as “in-place operation”
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Boettcher in view of Li, Lee, and Jones to incorporate the teachings of Li-2 wherein if the subsequent operation of the in-place operation is a subsequent in-place operation, changing (the second size of the output data of the in-place operation to the first size of the input data of the in-place operation and the output data of the in-place operation” to the output data corresponding to the output data of the subsequent in-place operation; and storing information to return to the second size, in association with the output data of the in-place operation. By doing so For each node, the inference pass will determine the range of root iterator variables first, using the shape of its output tensor. Then, Relay will figure out the leaf iterator variables by the relation of iterator variables and their relation recorded by stage node. Li-2 Page 13.
Jones teaches by pointing each pointer (a parameter PA1 may
be associated with nodes A and B of graph 100, PA2 with node C, and PA3 with nodes D, E, and F. In at least one embodiment, association between parameters and nodes of
a graph comprises a node's function referencing a corresponding parameter. [0053]) (a parameter comprises a reference to a memory region or buffer [0054])
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Boettcher in view of Li, Lee and Li-2 to incorporate the teachings of Jones wherein includes by pointing each pointer. By doing so some parameters are replaceable when efficacy of current optimizations is affected. Jones [0060].
Conclusion
9. The prior art made of record and not relied upon is considered pertinent to
applicant's disclosure.
Takemoto et al., (United States Patent Publication Number 20200019885) teaches “the processing returns to processing S102 to adjust a necessary parameter and feed back the necessary parameter to processing S104. In the example of FIG. 1, the parameter to be adjusted is A. Then, the parameter is optimized by repeating processing S104 and processing S105 until satisfactory strong classifier accuracy is obtained [0023]”
10. Any inquiry concerning this communication or earlier communications from the
examiner should be directed to Kweku Halm whose telephone number is (469) 295-
9144. The examiner can normally be reached on 7:30AM - 5:30PM Mon - Thur. If
attempts to reach the examiner by telephone are unsuccessful, the examiner's
supervisor, Sanjiv Shah can be reached on (571) 272-4098. 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 the
Patent Application Information Retrieval (PAIR) system. Status information for published
applications may be obtained from either Private PAIR or Public PAIR. Status information
for unpublished applications is available through Private PAIR only. For more
information about the PAIR system, see http://pair-direct.uspto.gov. Should you have
questions on access to the Private PAIR system, contact the Electronic Business Center
(EBC) at 866-217-9197 (toll-free).
/KWEKU WILLIAM HALM/Examiner, Art Unit 2166
/SANJIV SHAH/Supervisory Patent Examiner, Art Unit 2166
Prosecution Insights