DETAILED ACTION
Notice of Pre-AIA or AIA Status
1. The present application, filed on or after March 16, 2013, is being examined
under the first inventor to file provisions of the AIA .
Claim Interpretation
2. The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
3. The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitation(s) uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. Such claim limitation(s) is/are: “a computation unit” in claims 1-17.
Because this/these claim limitation(s) is/are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, it/they is/are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
If applicant does not intend to have this/these limitation(s) interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitation(s) recite(s) sufficient structure to perform the claimed function so as to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
Rejections - 35 USC § 112
4. The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.--The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
5. Claims 6-8 rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor, or for pre-AIA the applicant regards as the invention.
Regarding claims 6 and 7, the recitation of the phrases "such as" renders the claims indefinite. See MPEP § 2173.05(d).
Regarding claim 8, the recitation of the phrase “a training/configuring mode” renders the claim indefinite. It is unclear whether the slash “/” refers to “and” or “or" or “and/or”.
Therefore, the examiner comprehends the claims based on her best interpretations of these phrases.
Claim Rejections - 35 USC § 101
6. The following is a quotation of the appropriate paragraphs of 35 U.S.C. 101 that form the basis for the rejections under this section made in this Office action:
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.
7. Claims 1-20 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.
Under the 2019 PEG (now been incorporated into MPEP 2106), the revised procedure for determining whether a claim is "directed to" a judicial exception requires a two-prong inquiry into whether the claim recites: (1) any judicial exceptions, including certain groupings of abstract ideas (i.e., mathematical concepts, certain methods of organizing human interactions such as a fundamental economic practice, or mental processes); and (2) additional elements that integrate the judicial exception into a practical application (see MPEP § 2106.05(a)-(c), (e)-(h)).
Only if a claim (1) recites a judicial exception and (2) does not integrate that exception into a practical application, do we then look to whether the claim: (3) adds a specific limitation beyond the judicial exception that is not "well-understood, routine, conventional" in the field (see MPEP § 2106.0S(d)); or (4) simply appends well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception.
Claims 1-20 are directed to an abstract idea of variable accuracy computing.
Specifically, representative claim 1recites:
A computing system comprising:
a computation unit configured to receive a series of input data values and generate a series of output data values by performing operations on at least one received input data value and/or generated output data value;
an input to receive a tunable performance parameter separate to the series of input data values; and
(c) a controller, wherein the controller is configured, as a function of the received tunable performance parameter, to issue a control signal to the computation unit to control a level of accuracy of the operations and thereby affect a performance metric of the computation unit.
The claim limitations in the abstract idea have been highlighted in bold above; the remaining limitations are “additional elements”.
The highlighted portion of the claim constitutes an abstract idea under the 2019 Revised Patent Subject Matter Eligibility Guidance and the additional elements are NOT sufficient to amount to significantly more than the judicial exceptions, as analyzed below:
Step
Analysis
1. Statutory Category ?
Yes.
System/Apparatus
2A - Prong 1: Judicial Exception Recited?
Yes.
See the bolded portion as listed above.
Under its broadest reasonable interpretation (BRI), each or the combination of the limitations (a) and (c) recited in the bolded portion encompasses mental processes, i.e. data manipulation, evaluation and judgment, that can be performed in the human mind or by a human using a pen and paper.
Nothing in the bolded portion precludes the limitations (a) and (c) from practically being performed in the human mind with the aid of pen/paper or performed by a human manually.
As such, the bolded portion of instant claim 10, reciting a series of mental process, amounts to an abstract idea falling within the “Mental Process” grouping of Abstract Ideas defined by the 2019 PEG.
2A - Prong 2: Integrated into a Practical Application?
No.
Representative claim 1 recites the additional elements “a computing unit” and “a controller” at a high level of generality. Under the BRI, the claim requires merely to generally implement the abstract idea recited in the bolded portion using a general-purpose computer for performing activities in its ordinary capacity, such as data receiving/storing, transmitting and/or computing, via basic function of the computer. It is held that performing an abstract idea using a general-purpose computer system would not amount to significantly more than the abstract algorithm itself. See MPEP 2106.04(a)(2): if a claim limitation, under its broadest reasonable interpretation, covers mental processes except for the mention of generic computer components performing computing activities via basic function of the computer, then the claim is likely considered to be directed to an ineligible abstract idea, as it essentially describes a mental process that could be performed by a human without the computer components adding any significant practical application beyond the abstract concept itself.
Under the BRI, the additional (b) encompass a process of gathering the data/information necessary for performing the abstract idea. The claim does not require a particular device or sensor to receive the data information in a unique way that is never done before in the relevant technological field. See MPEP 2106.05(g)(3): … that were described as mere data gathering in conjunction with a law of nature or abstract idea. See also Mayo, 566 U.S. at 79, 101 USPQ2d at 1968; OIP Techs., Inc. v. Amazon.com, Inc., 788 F.3d 1359, 13863, 115 USPQ2d 1090, 1092-93 (Fed. Cir. 2015) (presenting offers and gathering statistics amounted to mere data gathering).
The claim as a whole does not meet any of the following criteria to integrate the abstract idea into a practical application:
An additional element reflects an improvement in the functioning of a computer, or an improvement to other technology or technical field;
an additional element that applies or uses a judicial exception to affect a particular treatment or prophylaxis for a disease or medical condition;
an additional element implements a judicial exception with, or uses a judicial exception in conjunction with, a particular machine or manufacture that is integral to the claim;
an additional element effects a transformation or reduction of a particular article to a different state or thing; and
an additional element applies or uses the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment, such that the claim as a whole is more than a drafting effort designed to monopolize the exception.
Various considerations are used to determine whether the additional elements are sufficient to integrate the abstract idea into a practical application. However, in all of these respects, the claim fails to recite additional elements which might possibly integrate the claim into a particular practical application. Instead, based on the above considerations, the claim would tend to monopolize the algorithm across a wide range of applications.
2B: Claim provides an Inventive Concept?
No.
Focusing on what the inventors have invented exactly, it is considered that the “core” of pending claim 1 is directed to an abstract idea of variable accuracy computing or an abstract algorithm of accuracy-oriented tunning. Under the BRI, all the claimed additional limitations recite activities that are all well-understood, routine, conventional in the art, they do not provide any inventive concepts or reflect a qualified improvement. The claim does not recite any additional element that can be treated as “significantly more”. See MPEP 2106.05.
The claim is therefore ineligible under 35 USC 101.
The dependent claims 2-17 inherit attributes of the independent claim 1, but do not add anything which would render the claimed invention a patent eligible application of the abstract idea. The claim merely extends (or narrows) the abstract idea which does not amount for "significant more" because it merely adds details to the algorithm which forms the abstract idea as discussed above.
Under the BRI, the additional limitations recited in claims 2-4 encompass insignificant extra-solution activities which are not qualified for meaningful limitations. At most, they only generally link the use of the judicial exception to a particular technological environment or field of use. Further, the additional limitations recited in claims 2-4 do not reflect any “inventive concept” under the 2019 PEG.
Claims 18-20 are rejected under 35 U.S.C. § 101 for the same reason as for claims 1-17 set forth above.
Claim Rejections - 35 USC § 102
8. The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention; or
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
9. Claims 1, 3, 5-12, and 17-19 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Kemp et al. (ASPDAC '21: Proceedings of the 26th Asia and South Pacific Design Automation Conference, Pages 248 – 253, 29 January 2021).
Regarding claims 1 and 18, Kemp discloses a computing system and a method of operating the computing system (pages 249-250, Section 4: PROPOSED APPROACH), the system comprising: a computation unit (e.g., the IM circuit in Fig. 3 for performing the iterative approximate computing; see also the discussion of the “Application” module of MIPAC in Fig. 2(b)) configured to receive a series of input data values (e.g., the “input” shown in Fig. 3) and generate a series of output data values by performing operations on at least one received input data value and/or generated output data value (page 249, left column, 2nd paragraph; see also the discussion of Fig. 3); an input to receive a tunable performance parameter (e.g., the “Target accuracy” as shown in Fig. 3) separate to the series of input data values (page 249, left column, 3rd paragraph: “We propose the design of a lightweight quality controller which determines the optimal NOI for a given input to achieve optimal energy-accuracy trade-off providing fine-grained dynamic accuracy configuration”; page 250, left column, 1st paragraph: “the optimal NOI can be derived from the input and/or some input-dependent signals from the IM circuit”; page 250, left column, 2nd to last paragraph: “Based on the target accuracy 𝛼, the output of the quality predictor that corresponds to the target accuracy is selected”); and a controller (MIPAC in Figs. 2(b) and 3), wherein the controller is configured, as a function (e.g., generating the optimal number of iteration “NOI”) of the received tunable performance parameter, to issue a control signal to the computation unit to control a level of accuracy of the operations and thereby affect a performance metric (e.g., the total power constrain 𝛽) of the computation unit (page 249, left column, 2nd paragraph: “a lightweight quality controller which determines the optimal NOI for a given input to achieve optimal energy-accuracy trade-off providing fine grained dynamic accuracy configuration”; see also discussion of MIPAC and/or power capping in Figs. 3 and 4).
Regarding claim 3, Kemp discloses: wherein the computation unit comprises a generative pre-trained transformer (page 248, left column, last paragraph; see also pages 249-250, Sections 4, discussion of the “IM” circuitry).
Regarding claim 5, Kemp discloses: the computing system according to claim 1, configured to generate a composite tunable performance parameter (e.g., the “𝑛 quality predictors” from which the quality predictor, e.g., 𝑃𝛼1, that produces the minimum NOI for a given input of the target accuracy 𝛼1 is selected) based at least in part on the received tunable performance parameter (page 250, left column, 2nd to last paragraph), optionally wherein the controller is configured to issue the control signal as a function of the composite tunable performance parameter (see discussion of the “quality predictor” in Sections 5.1, 5.2 and 6.1).
Regarding claim 6, Kemp discloses: configured to adjust (e.g., select the best one from the “𝑛 quality predictors”) a value of the composite tunable performance parameter based on at least one received input data value (page 250, left column, last paragraph; page 253, left column, 2nd to last paragraph; see also discussion of the “quality predictor” and/or “quality prediction” in Sections 4, 5.1, 5.2 and 6.1).
Regarding claim 7, Kemp discloses: wherein said function is a function of the received input data value (page 249, left column, 2nd paragraph; see also discussion of MIPAC in Figs. 3 and 4).
Regarding claim 8, Kemp discloses: wherein the controller is configured, in a configuring mode (e.g., Fig. 4), to: vary a value of the control signal (see discussion of power capping in Fig. 4) and define said function (generating the predictor set from which the optimal number of iteration “NOI” is selected) based on an effect of varying the value of the control signal on the performance metric (pages 251-252, Section 5.2 discusses power capping to ensure that the power overhead of the quality controller meets the power constraint 𝛽).
Regarding claim 9, Kemp discloses: wherein the controller comprises a look-up table and is configured to obtain a value of the control signal by accessing the look-up table based at least on a value of the tunable performance parameter (page 250, left column, last paragraph – right column, 1st paragraph; see also discussion of Fig. 5).
Regarding claim 10, Kemp discloses: wherein the performance metric comprises a power consumption of the computing system (see discussion of the power constraint 𝛽 in Sections 5.1 and 5.2).
Regarding claim 11, Kemp discloses: wherein in response to the control signal the computation unit is operative to adjust values (e.g., number of iterative approximate computing) derived therefrom that are used by the computation unit in performing said operations (page 249, left column, 2nd paragraph; page 250, left column, 1sr and 2nd paragraphs).
Regarding claim 12, Kemp discloses: wherein the computation unit is operative to prevent use of values (e.g., the values of NOI that do not comply with the target accuracy requirements and power constraint) derived therefrom from being used by the computation unit in performing said operations (see, e.g., pages 249-250, Sections 3 and 4).
Regarding claim 17, Kemp discloses an integrated circuit or a device (e.g., Fig. 3), comprising a computing system (the MIPAC-enabled IM circuit including the proposed lightweight quality controller) according to claim 1 (see pages 249-250, Section 4: PROPOSED APPROACH).
Regarding claim 19, Kemp discloses a method of configuring a controller of a computing system comprising a computation unit (IM circuit in Fig. 3) and said controller (MIPAC in Fig. 3), the computation unit configured to receive a series of input data values and generate a series of output data values by performing operations on at least one received input data value and/or generated output data value, the controller configured, as a function of a tunable performance parameter, to issue a control signal to the computation unit to control a level of accuracy of the operations and thereby affect a performance metric of the computation unit (see discussion of claims 1 and 18 above), the method comprising: varying a value of the tunable performance parameter and defining said function based on an effect of varying the value of the tunable performance parameter on the performance metric (see discussion of claim 8 above).
Claim Rejections - 35 USC § 103
10. 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.
11. Claims 2, 4 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Kemp et al. in view of Clemons et al. (US 20230111375 A1).
Regarding claim 2, Kemp discloses: wherein the computation unit comprises application-specific approximate computing for machine learning circuitry (Abstract; see also pages 249-250, Sections 4, discussion of the “IM” circuitry).
Kemp is silent on: said machine learning circuitry is digital and/or analog neural network circuitry.
Clemons discloses a controller (105 in Figs. 1A and 1C) for controlling dynamic configuration and execution of an augmented digital and/or analog neural network circuitry (110 in Figs. 1A and 1C) in real-time according to performance constraints (para. 0041: “application of the configuration selection dynamically transforms the augmented neural network model 110 into a variant of the original neural network model associated with inference latency, inference throughput, energy consumption, and/or accuracy characteristics adapted to satisfy real-time performance constraints”; see also discussion of Fig. 1B).
In view of Clemons, it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to apply Kemp’s system/method of “Dynamic Input-Aware Accuracy Control for Dynamic Auto-Tuning of Iterative Approximate Computing” to a digital and/or analog neural network circuitry for machine learning computing. Doing so would allow, through dynamic configuration, the augmented neural network model may adapt to real-time changes in the performance constraints while the trained weights for an original (before augmentation) neural network model may be used without modification (Clemons, Abstract). It has been held that the mere application of a known technique to a specific instance by those skilled in the art would have been obvious.
Regarding claim 4, Kemp is silent on: wherein the computation unit is configured to operate as a Large Language Model (LLM), wherein the input data values are, or are part of, input tokens of the LLM and wherein the output data values are, or are part of, output tokens of the LLM.
The teaching of Clemons includes: wherein the computation unit (e.g., PPU 400 in Fig. 4) is configured to operate as a Large Language Model (LLM) (para. 0057: “a processor such as the PPU 400 may be configured to implement a neural network model, such as the augmented neural network model 110 and/or 300”; para. 0059: “The PPU 400 may be configured to accelerate numerous deep learning systems and applications for … real-time language translation”), wherein the input data values are, or are part of, input tokens of the LLM (see discussion of “Input Tensor” in Figs. 1A and 1C) and wherein the output data values are, or are part of, output tokens of the LLM (see discussion of “Output” from the augmented neural network model in Figs. 1A and 1C; see also para. 0115).
It would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate Kemp’s technique of accuracy-oriented tunning into Clemons’ dynamic configuration and execution of the augmented LLM to arrive the claimed invention. Doing so would allow, through dynamic configuration, the augmented LLM model may adapt to real-time changes in the performance constraints while the trained weights for an original (before augmentation) neural network model may be used without modification (Clemons, Abstract). It has been held that the mere application of a known technique to a specific instance by those skilled in the art would have been obvious.
Regarding claim 20, Kemp is silent on: a non-transitory computer readable storage medium having a computer program stored thereon which, when executed on one or more processors of a computing system, causes the computing system to carry out the method of claim 18.
Clemons discloses a non-transitory computer readable storage medium having a computer program stored thereon which, when executed on one or more processors of a computing system (e.g., 100 in Fig. 1A), causes the computing system to carry out (para. 0025: “various functions may be carried out by a processor executing instructions stored in memory…”; para. 0028: “various functions may be carried out by a processor executing instructions stored in memory. The method may also be embodied as computer-usable instructions stored on computer storage media”): receiving, at a computation unit (110) of a computing system, a series of input data values (e.g., “Input Tensor” Fig. 1A); receiving, at an input of the computing system, a tunable performance parameter (e.g., “performance constrains” Fig. 1A) separate to the series of input data values, generating, at the computation unit, a series of output data values (e.g., “Output” from 110 in Fig. 1A) by performing operations on at least one received input data value and/or generated output data value; and controlling, by a controller (105 in Fig. 1A) of the computing system and as a function of the received tunable performance parameter, a level of performance of the operations and thereby affecting a performance metric of the computation unit (para. 0003).
It would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to incorporate Clemons’s teaching of the non-transitory computer readable storage medium and the computer program into the Kemp system/method to arrive the claimed invention. A persons of ordinary skill in the art would conceive and apply such a computing program implementation to the Kemp system/method without needing inventive skill but depending on practical considerations and according to the dictates of the circumstances (Clemons, para. 0028: “persons of ordinary skill in the art will understand that any system that performs method 120 is within the scope and spirit of embodiments of the present disclosure”).
12. Claims 13-16 are rejected under 35 U.S.C. 103 as being unpatentable over Kemp et al. in view of FRACHTENBERG et al. (US 20250181894 A1).
Regarding claim 13, Kemp is silent on: wherein said operations comprise applying a weight value to at least one received input data value or value derived therefrom and wherein the computation unit is operative to adjust a number of bits of the weight value used by the computation unit in performing said operations.
FRACHTENBERG discloses systems and methods for dynamically adjusting amounts of precision (i.e., numbers of bits) used to represent and store individual weights of a neural network model during machine learning operations (Abstract), comprising: a computation unit (102 in Fig. 1A) configured to receive a series of input data values and generate a series of output data values (para. 0023-0024) by performing said operations on at least one received input data value and/or generated output data value (para. 0038), wherein said operations comprise applying a weight value to at least one received input data value or value derived therefrom and wherein the computation unit is operative to adjust a number of bits of the weight value used by the computation unit in performing said operations (para. 0016, 0019, 0033-0040).
Since Kemp teaches the general conditions of the computation unit for performing iterative computing operations such as machine learning (Abstract), it would have been obvious to one having ordinary skill in the art before the effective filing date of the claimed invention to apply Kemp’s technique of accuracy-oriented tunning to FRACHTENBERG’s computation unit to arrive the claimed invention, as an intended use of known technologies. Doing so would allow for improving energy efficiency and performance by allowing for slight errors in intermediate computations (Kemp, Abstract). It has been held that the mere application of known technologies to a specific instance by those skilled in the art would have been obvious.
Regarding claim 14, Kemp does not but FRACHTENBERG teaches: wherein the computation unit is operative to prevent or disable use of one or more least significant bits of the weight value from being used by the computation unit in performing said operations (para. 0013: “using lower precision (i.e., a lower number of memory bits) to represent weights of the neural network model … In other words, the amount of memory hardware required to store neural network model weights can be reduced by storing/representing neural network model weights with lower precision”; see also para. 0019, 0031, 0034, 0038, 0040). As such, the combination of Kemp/FRACHTENBERG renders the claimed invention obvious.
Regarding claim 15, Kemp does not but FRACHTENBERG teaches: wherein said operations comprise accumulating a plurality of received input data values or data values derived therefrom to generate accumulated data values (para. 0002: “(b) one or more hidden layers of neurons that receive weighted outputs from the input layer (or in the case of multiple hidden layers, …”; para. 0011: “DNN models can be enormous, sometimes comprising billions of weights”; para. 0012: “ the weights of a DNN model are stored across many physical computing units …”), and wherein the computation unit is operative to adjust a number of bits of the data values which are subject to the accumulation and/or of the accumulated data values (para. 0016, 0019, 0033-0040). As such, the combination of Kemp/FRACHTENBERG renders the claimed invention obvious.
Regarding claim 16, Kemp does not but FRACHTENBERG teaches: wherein the computation unit is operative to prevent or disable use of one or more least significant bits of the data values which are subject to the accumulation and/or of the accumulated data values (para. 0013, 0019, 0031, 0034, 0038, 0040). As such, the combination of Kemp/FRACHTENBERG renders the claimed invention obvious.
Citation of Relevant Prior Art
13. The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Menon et al., Auto-tuning Parameter Choices in HPC Applications using Bayesian Optimization, 2020 IEEE International Parallel and Distributed Processing Symposium (IPDPS)
BONDALAPATI et al., Reconfigurable Computing Systems, PROCEEDINGS OF THE IEEE, VOL. 90, NO. 7, JULY 2002
Contact Information
14. Any inquiry concerning this communication or earlier communications from the examiner should be directed to XIUQIN SUN whose telephone number is (571)272-2280. The examiner can normally be reached 9:30am-6:00pm.
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://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Shelby A. Turner can be reached on (571) 272-6334. 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.
/X.S/Examiner, Art Unit 2857
/MOHAMMAD K ISLAM/Primary Examiner, Art Unit 2857
Prosecution Insights