International Business Machines CorporationDownload PDFPatent Trials and Appeals BoardSep 14, 20212021000266 (P.T.A.B. Sep. 14, 2021) Copy Citation UNITED STATES PATENT AND TRADEMARK OFFICE UNITED STATES DEPARTMENT OF COMMERCE United States Patent and Trademark Office Address: COMMISSIONER FOR PATENTS P.O. Box 1450 Alexandria, Virginia 22313-1450 www.uspto.gov APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO. 16/115,828 08/29/2018 Ali Javadiabhari P201805390US01 9533 108686 7590 09/14/2021 IBM IPLAW (GLF) c/o Garg Law Firm, PLLC 800 Bonaventure Way, suite 115 Sugar Land, TX 77479 EXAMINER DOAN, NGHIA M ART UNIT PAPER NUMBER 2851 NOTIFICATION DATE DELIVERY MODE 09/14/2021 ELECTRONIC Please find below and/or attached an Office communication concerning this application or proceeding. The time period for reply, if any, is set in the attached communication. Notice of the Office communication was sent electronically on above-indicated "Notification Date" to the following e-mail address(es): dpandya@garglaw.com garglaw@gmail.com uspto@garglaw.com PTOL-90A (Rev. 04/07) UNITED STATES PATENT AND TRADEMARK OFFICE _________ BEFORE THE PATENT TRIAL AND APPEAL BOARD __________ Ex parte ALI JAVADIABHARI, JAY M. GAMBETTA, ANDREW W. CROSS, LUCIANO BELLO, and MARCO PISTOIA __________ Appeal 2021-000266 Application 16/115,828 Technology Center 2800 ___________ Before ADRIENE LEPIANE HANLON, DONNA M. PRAISS, and SHELDON M. McGEE, Administrative Patent Judges. HANLON, Administrative Patent Judge. DECISION ON APPEAL A. STATEMENT OF THE CASE The Appellant1 filed an appeal under 35 U.S.C. § 134(a) from an Examiner’s decision finally rejecting claims 1–6 and 8–20. We have jurisdiction under 35 U.S.C. § 6(b). We REVERSE. 1 “Appellant” refers to “applicant” as defined in 37 C.F.R. § 1.42. The Appellant identifies the real party in interest as International Business Machines Corporation. Appeal Brief dated June 24, 2020 (“Appeal Br.”), at 2. Appeal 2021-000266 Application 16/115,828 2 Representative claim 1 is reproduced below from the Claims Appendix to the Appeal Brief. The claim limitation at issue is italicized.2 1. A method for design optimization of a quantum circuit, comprising: analyzing a first quantum circuit design based on at least one of a set of design criteria, wherein the first quantum circuit design comprises a set of quantum logic gates, and wherein a design criterion in the set of design criteria comprises changing a size of a matrix of transformations corresponding to a number of qubits employed in the first quantum circuit design, wherein analyzing the first quantum circuit design further comprises determining a first entanglement state of a first qubit; and modifying the first quantum circuit design into a transformed quantum circuit design, the modifying causing the transformed quantum circuit design to perform an operation implemented in the first quantum circuit design with a changed matrix of transformations. Appeal Br. 14. The Examiner maintains the following grounds of rejection on appeal: (1) claims 1–4, 8–16, 19, and 20 under 35 U.S.C. § 102(a)(1) as anticipated by Williams;3 and (2) claims 5, 6, 17, and 18 under 35 U.S.C. § 103 as unpatentable over Williams in view of Mosca.4 B. DISCUSSION The Examiner finds Williams discloses a method for design optimization of a quantum circuit comprising, inter alia, the analyzing step recited in claim 1 “wherein analyzing the first quantum circuit design further comprises determining 2 The limitation at issue in claim 1 is also recited in independent claims 11 and 20. See Appeal Br. 15, 17. 3 US 2006/0123363 A1 to Williams et al., published June 8, 2006 (“Williams”). 4 US 2018/0039903 A1 to Mosca et al., published February 8, 2018 (“Mosca”). Appeal 2021-000266 Application 16/115,828 3 a first entanglement state of a first qubit” (underlining omitted; italics added). Final Act. 4–55 (citing Williams ¶¶ 97, 99, 125, 126; Williams Figs. 5A, 5B). At the outset, it is necessary to interpret the terms “analyzing” and “determining” in the claim limitation at issue. See In re Hiniker Co., 150 F.3d 1362, 1369 (Fed. Cir 1998) (“the name of the game is the claim”). The Examiner finds that the terms “analyze,” “evaluate,” and “determine” are synonyms for the term “compute” and then interprets the term “determining” in claim 1 to mean “computing.” Ans. 5–6;6 see also Reply Br. 27 (contending that the “Examiner appears to be first interpreting determining a first entanglement state of a first qubit as computing a first entanglement state of a first qubit, then providing support in the reference for this interpretation” (original emphasis omitted; italics added)). The Appellant, however, argues that “[c]laim 1 does not recite ‘compute’” and “when interpreting claim 1, synonyms for terms not present in the claim – such as ‘compute’ . . . – should not be adopted.” Reply Br. 2. The Appellant does not define the terms “analyzing” or “determining” in the Specification and neither the Appellant nor the Examiner provides a definition for either term on appeal. Nonetheless, the Appellant discloses that “an analysis pass of an embodiment collects information about the quantum circuit.” Spec. ¶ 14 (emphasis added). More specifically, the Appellant discloses: Analysis pass component 410 analyzes the input quantum circuit and collects information, such as a number of quantum logic gates, the types of quantum logic gates, the depth of the quantum circuit, and the number of qubits. Analysis pass component 410 includes a set of analysis passes capable of being executed by pass manager component 408. For example, analysis pass component 410 5 Final Office Action dated March 27, 2020. 6 Examiner’s Answer dated September 14, 2020. 7 Reply Brief dated October 8, 2020. Appeal 2021-000266 Application 16/115,828 4 analyzes a first quantum circuit design based on at least one of a set of design criteria for a new quantum circuit design. The quantum circuit design of an embodiment includes a set of quantum logic gates. A design criterion of the set of design criteria of an embodiment includes changing a size of a matrix of transformations corresponding to a number of qubits employed in the quantum circuit design. Once the analysis pass component 410 finishes an analysis pass on the quantum circuit, the collected information is returned to the pass manager component 408. The pass manager component 408 updates information about the input quantum circuit design and stores the information as an updated quantum circuit design. Spec. ¶ 70 (emphasis added). Based on the Appellant’s disclosure, we interpret the term “analyzing,” as used in the claims on appeal, to mean “collecting information.”8 Our interpretation is consistent with the ordinary meaning of the term “analyze.” See Definition of “analyze,” dictionary.cambridge.org/us/dictionary/english/analyze (last visited Sept. 9, 2021) (defining “analyze” as “to study or examine something in detail, in order to discover more about it” (emphasis added)). Moreover, “determining a first entanglement state of a first qubit” is part of the analyzing step recited in claim 1. Appeal Br. 14 (emphasis added); see also Spec. ¶ 18 (disclosing that “analyzing the first quantum circuit design determines a first entanglement state of a first qubit” (emphasis added)). Thus, consistent with our interpretation of the term “analyze,” we interpret the term “determine” to mean “to discover 8 See In re Morris, 127 F.3d 1048, 1054 (Fed. Cir. 1997) (During examination, “the PTO applies to the verbiage of the proposed claims the broadest reasonable meaning of the words in their ordinary usage as they would be understood by one of ordinary skill in the art, taking into account whatever enlightenment by way of definitions or otherwise that may be afforded by the written description contained in the applicant’s specification.”). Appeal 2021-000266 Application 16/115,828 5 the facts . . . about something” (emphasis added).9 Definition of “determine,” dictionary.cambridge.org/us/dictionary/english/determine (last visited Sept. 9, 2021) (defining “determine” as “to discover the facts or truth about something”). Claim 1 recites, inter alia, the step of “analyzing a first quantum circuit design . . . wherein analyzing the first quantum circuit design further comprises determining a first entanglement state of a first qubit.” Appeal Br. 14 (emphasis added). Based on our interpretation of the terms “analyzing” and “determine” identified above, we interpret the claim limitation at issue to mean collecting information about or discovering, not calculating, a first entanglement state of a first qubit.10 See Reply Br. 3 (contending that “determining is part of an analysis of a property of a quantum circuit design – in other words, a measurement operation, not a computation” (original emphasis omitted; italics added)). Next, we turn to the anticipation rejection on appeal. “To anticipate a claim, a prior art reference must disclose every limitation of the claimed invention, either explicitly or inherently.” In re Schreiber, 128 F.3d 1473, 1477 (Fed. Cir. 1997); see also In re Oelrich, 666 F.2d 578, 581 (CCPA 1981) (“Inherency . . . may not be established by probabilities or possibilities. The mere fact that a certain thing may result from a given set of circumstances is not sufficient.” (quoting Hansgirg v. Kemmer, 102 F.2d 212, 214 (CCPA 1939))). 9 In contrast to the terms “analyze” and “determine,” the term “compute” means “to calculate an answer or amount by using a machine” (emphasis added). Definition of “compute,” dictionary.cambridge.org/us/dictionary/english/compute (last visited Sept. 9, 2021). 10 We note that the Specification does not disclose how an entanglement state of a qubit is determined. Appeal 2021-000266 Application 16/115,828 6 To support the anticipation rejection on appeal, the Examiner relies on paragraphs 125 and 126 of Williams which disclose a method of data entry. Final Act. 5. Williams discloses, in relevant part: (1) Step 1: Make the Bit String-Eigenstate Correspondence: Given a string of 2n classical bits, place these in 1-to-1 correspondence with the set of 2n n-bit eigenstates. . . . (2) Step 2: Map Bits to Superposition[11] State: Construct an equally weighted superposition that is peaked at only those eigenstates which correspond to a “1” in the given bit string sequence (i.e. the classical data). . . . In general, this will be an entangled state[12] of n quantum bits that encodes a specific sequence of 2n classical bits. . . . (3) Step 3: Compute Unitary Transformation Needed to Make this Superposition State Starting From (Say) ǀ00 . . . 0> . . . . (4) Step 4: Compute the Quantum Circuit that Implements this Unitary Transformation . . . . Williams ¶¶ 124–126, 128 (emphasis added). The Examiner appears to rely on the combination of encoding step (2) and computing steps (3) and (4) to illustrate an analysis of a first entanglement state of a first qubit as recited in claim 1. Ans. 7. The Appellant argues: Williams encodes a string of 2ʺ classical bits into an entangled state of n quantum bits, computes a unitary transformation needed to make the entangled state, then computes a quantum circuit implementing the unitary transformation. Thus, the known entangled state is the output of the quantum circuit, rather than a result of analyzing the circuit. In addition, because in Williams the entanglement state is encoded first, then a design generated to result in the entanglement state, Williams 11 According to the Appellant, “[s]uperposition means that each qubit can represent both a 1 and a 0 at the same time.” Spec. ¶ 6. 12 According to the Appellant, “[e]ntanglement means that qubits in a superposition can be correlated with each other in a non-classical way; that is, the state of one (whether it is a 1 or a 0 or both) can depend on the state of another.” Spec. ¶ 6. Appeal 2021-000266 Application 16/115,828 7 has no need to analyze a circuit to determine an entanglement state that is already known. Appeal Br. 11 (original emphasis omitted; italics added); see also Reply Br. 3. As discussed above, we interpret the claim limitation at issue to mean collecting information about, not calculating, a first entanglement state of a first qubit. This portion of Williams, however, does not describe collecting information about an existing entanglement state as recited in claim 1. Rather, as the Appellant contends, Williams constructs an entangled state of n quantum bits that encodes a specific sequence of 2n classical bits, computes a unitary transformation needed to make the entangled state, and then computes a quantum circuit that implements the unitary transformation. Williams ¶¶ 125, 126, 128; Appeal Br. 11. The Examiner also relies on paragraphs 97 and 99 of Williams to support the anticipation rejection on appeal. Final Act. 5. Williams discloses, in relevant part: Permutation matrices are a special class of 4ⅹ4 unitary matrices possessing a single 1 in each row and column. Of the 4!=24 possible permutation matrices, 16 are maximally entangling operations and 8 are non-entangling operations. In FIGS. 5A and 5B, we show the optimal quantum circuit for each maximally-entangling permutation matrix. . . . The optimal circuits in FIGS. 5A and 5B were found using a numerical technique that incrementally explores circuit templates of increasing complexity until the simplest sufficient template is found. This is feasible for 2-qubit gates but not, in general for n-qubit gates, due to the exponentially growing number of possible templates to try. Similarly, the invention can compute optimal circuits for the remaining 8 non-entangling operations as shown in FIG. 6. We first compute the entangling power of the unitary operator. If this is zero this suggests that the unitary operator is either a direct product of 1- qubit gates or equivalent to a SWAP gate up to 1-qubit gates. Williams ¶¶ 97–99 (emphasis added). Appeal 2021-000266 Application 16/115,828 8 Williams, in paragraphs 97–99, does not expressly describe collecting information about a first entanglement state of a first qubit as recited in claim 1. Nonetheless, the Examiner contends that the phrase “‘incrementally exploring circuit templates’ [in paragraph 98 of Williams] . . . means analyzing . . . these quantum circuits until the simplest/optimal quantum circuit is found” and “involves the determination . . . of a first entanglement states [sic] of a first qubit associated with the circuit templates” as recited in claim 1. Ans. 8 (emphasis added). Williams’ Figures 5A and 5B illustrate an optimal quantum circuit for each maximally-entangling permutation matrix referred to in paragraph 97 of Williams. See Williams ¶ 23. Williams discloses that the optimal circuits “were found using a numerical technique that incrementally explores circuit templates of increasing complexity until the simplest sufficient template is found.” Williams ¶ 98 (emphasis added). Significantly, the Examiner does not explain whether the numerical technique described in Williams collects information about an entanglement state of a qubit as claimed or is merely a computation method. See Reply Br. 4 (arguing that “Williams starts with a matrix known to be maximally- entangling, then generates circuit templates of increasing complexity implementing the matrix” (original emphasis omitted; italics added)). Therefore, the Examiner has failed to show that paragraphs 97–99 of Williams describe the claim limitation at issue. In sum, the Examiner has not established, by a preponderance of the evidence, that Williams either expressly or inherently describes an analyzing step comprising “determining a first entanglement state of a first qubit” as recited in the independent claims on appeal. See In re Oetiker, 977 F.2d 1443, 1445 (Fed. Cir. 1992) (the examiner bears the initial burden of presenting a prima facie case of Appeal 2021-000266 Application 16/115,828 9 unpatentability). Therefore, the anticipation rejection of claims 1–4, 8–16, 19, and 20 based on Williams is not sustained. The Examiner’s reliance on Mosca does not cure the deficiencies in Williams. Final Act. 7; see also Appeal Br. 12 (arguing that the “Examiner combines Mosca with Williams only to allege teachings of specific features of claims 5–6 and 17–18” (emphasis omitted)). Therefore, the obviousness rejection of claims 5, 6, 17, and 18 based on the combination of Williams and Mosca also is not sustained. C. DECISION The Examiner’s decision is reversed. In summary: Claims Rejected 35 U.S.C. § Reference(s)/Basis Affirmed Reversed 1–4, 8–16, 19, 20 102(a)(1) Williams 1–4, 8–16, 19, 20 5, 6, 17, 18 103 Williams, Mosca 5, 6, 17, 18 Overall Outcome 1–6, 8–20 REVERSED Copy with citationCopy as parenthetical citation