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 .
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claims 1-7 and 9-18 have been examined.
Response to Arguments
Applicant's arguments filed on 5/7/2026 have been fully considered but they are not persuasive.
Regarding Applicant’s remarks, Applicant mainly argues that the prior art of record does not explicitly disclose “determining whether the number of GPU requests set by the user is physically satisfiable; and when it is determined that the number of GPU requests is physically unsatisfiable, identifying a pre-set user policy, and when it is determined that multi-node allocation is allowed according to a result of the identifying, allocating a GPU over multiple nodes to satisfy the number of GPU requests.” Specifically, Applicant argues that Sivanthanu merely discloses that GPUs are allocated across multiple servers to satisfy a large job, and Sivanthanu does not disclose conditions required to achieve user-policy-controlled fallback that respects the user’s operational preference. However, the examiner disagrees.
In response to applicant's argument that the references fail to show certain features of the invention, it is noted that the features upon which applicant relies (i.e. user-policy-controlled fallback that respects the user’s operational preference) are not recited in the rejected claim(s). Although the claims are interpreted in light of the specification, limitations from the specification are not read into the claims. See In re Van Geuns, 988 F.2d 1181, 26 USPQ2d 1057 (Fed. Cir. 1993).
Furthermore, based on broadest reasonable interpretation consistent with the Specification, “pre-set user policy” may be instructions configured by administrators of the auto-scaler to allow automatic scheduling based on GPU metrics as disclosed by Sivanthanu. Without further clarifications regarding the context and content of the pre-set user policy, the combination of references reasonably teaches or at least suggests the disputed limitations. Therefore, Applicant’s argument is not persuasive in light of above explanation.
Regarding 35 U.S.C. 112(f) interpretation, the interpretation is withdrawn in light of Amendment filed on 5/7/26.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1, 2, 9, 10, 11, 12 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Sivanthanu U.S. 2022/0318052 in view of Ni U.S. 2023/0109368 (hereinafter Ni).
As per claim 1, 10 and 11, Sivanthanu discloses a cloud management method/computer readable recording medium/system comprising:
a processor configured to:
collect data for allocating GPU resources in a large-scale container operating environment (Sivanthanu: [0028]: monitor workloads that are currently running and hardware capacity that is currently available anywhere around the world in the cloud for scheduling GPU services); and
generate a multi-metric based on the collected data, to set a scheduling priority for the generated pod (Sivathanu: [0028]: scale up a job, i.e. create new pod by monitoring GPU utilization/collected data); and
perform a scheduling operation for allocating GPU resources according to the set scheduling priority (Sivanthanu: [0028]-[0029]: scheduler allocates resources based on utilization…scheduling priority; [0066]: new workload may be associated with a higher priority or tier than the current workload).
Sivanthanu discloses scale up or down a job based by monitoring/tracking workload, including pods (Sivanthanu: [0028]: monitor workloads to scale up or down a job; [0058]: preparing schedules corresponding to workloads include jobs, model, and/or pods);
wherein the performing the scheduling operation comprises:
determining whether the number of GPU requests set by the user is physically satisfiable; and
when it is determined that the number of GPU requests is physically unsatisfiable, identifying a pre-set user policy, and when it is determined that multi-node allocation is allowed according to a result of the identifying, allocating a GPU over multiple nodes to satisfy the number of GPU requests (Sivathanu: [0099]-[0101]; [0114]: GPUs are allocated across multiple servers for a large job).
Sivanthanu does not explicitly disclose generating new pod based on the multi-metric. However, Ni discloses autoscale number of pods based on GPU metric (Ni: [0034]-[0035]: autoscale number of pods based on GPU metrics). It would have been obvious to one having ordinary skill in the art to generate new pods based on GPU metrics in the process of scaling up AI workload because generating new pods is well known in the art for Kubernetes systems.
As per claim 2 and 12, Sivanthanu as modified discloses the limitations of claims 1 and 11 respectively. Sivanthanu further discloses wherein the step of setting the scheduling priority comprises, when a new pod is generated, setting a scheduling priority for the generated pod by reflecting a priority set by a user and a number of times of trying rescheduling (Sivathanu: [0028]: scale up a job, i.e. generate new pod; [0033]: establish priority; [0110]-[0112]: priority-based scheduling depends on pass value, i.e. how many rounds or number of times of rescheduling).
As per claim 9 and 18, Sivanthanu as modified discloses the limitations of claims 1 and 11 respectively. Sivanthanu as modified further discloses wherein the step of collecting data comprises collecting GPU resources comprising GPU utilization, GPU memory, GPU clock, GPU architecture, GPU core, GPU power, GPUtemperature, GPU process resource, GPU NVlink pair, GPU return, and GPU assignment (Sivathanu: [0028]: scheduler monitors utilization of GPU utilization to collect data; [0066]: scale up or down based on resource utilization; Ni: [0034]-[0035]).
Claims 3-5 and 13-15 are rejected under 35 U.S.C. 103 as being unpatentable over Sivanthanu in view of Ni and further in view of Zhang et al. U.S. 2022/0291956 (hereinafter Zhang).
As per claim 3 and 13, Sivanthanu as modified discloses the limitations of claims 1 and 11 respectively. Sivanthanu does not explicitly disclose wherein performing the scheduling operation comprises, when performing the scheduling operation, performing a node filtering operation, a GPU filtering operation, a node scoring operation, and a GPU scoring operation. However, Zhang discloses filtering and scoring nodes and GPUs according to requests (Zhang: [0022]-[0029]; [0048]: multiple tasks may share the GPUs based on utilization of GPU resources in the distributed container cluster). It would have been obvious one having ordinary skill in the art to filter node and score GPUs to ensure the load balance of nodes in the cluster, enhance the utilization of GPU resources in the distributed container cluster, better meet the scheduling requirements, and allow containers to complete tasks faster.
As per claim 4 and 14, Sivanthanu as modified discloses the limitations of claims 3 and 13 respectively. Sivanthanu as modified further discloses wherein the step of performing the scheduling operation comprises, when performing the GPU filtering operation and the GPU scoring operation, reflecting a number of GPU requests set by a user and a requested GPU memory capacity (Zhang: [0022]-[0029]; [0048]). Same rationale applies here as above in rejecting claim 3.
As per claim 5 and 15, Sivanthanu as modified discloses the limitations of claims 4 and 14 respectively. Sivanthanu as modified further discloses wherein the step of performing the scheduling operation comprises: when it is determined that the number of GPU requests is physically satisfiable, performing a GPU filtering operation and a GPU scoring operation with respect to an available GPU; and allocating GPU resources based on a result of the GPU filtering operation and the GPU scoring operation (Zhang: [0022]-[0029]: select optimal nodes based on GPU scores; [0048]: scheduling containers to the most adaptive node based on the metric state, free memory and allocation of graphics cards at the node). Same rationale applies here as above in rejecting claim 3.
Claims 6-7 and 16-17 are rejected under 35 U.S.C. 103 as being unpatentable over Sivanthanu in view of Ni and further in view of Zhang and further in view of Garg et al. U.S. 2021/0011773 (hereinafter Garg).
As per claim 6 and 16, Sivanthanu as modified discloses the limitations claims 5 and 15 respectively. Sivanthanu does not explicitly disclose wherein the step of performing the scheduling operation comprises, when it is determined that a total number of GPU requests set for a plurality of pods, respectively, is physically unsatisfiable, identifying a partitionable GPU memory, partitioning one GPU memory into a plurality of GPU memories, and allocating the plurality of partitioned GPU memories to a plurality of pods to allow the plurality of pods to share one physical GPU device. However, Garg teaches or at least suggests the limitations (Garg: [0015]: partitioning GPU memory to support vGPU profiles associated with different workloads/vms). It would have been obvious to one having ordinary skill in the art partition GPU memories for multiple instances to fully utilize capacity of GPUs as well known in the scheduling process.
As per claim 7 and 17, Sivanthanu as modified discloses the limitations of claims 5 and 15 respectively. Sivanthanu as modified further discloses wherein the step of performing the scheduling operation comprises, when it is determined that the number of GPU requests is physically unsatisfiable, identifying a partitionable GPU memory, partitioning one GPU memory into a plurality of GPU memories, and allocating a part or all of the plurality of partitioned GPU memories to the pod (Garg: [0015]: partition GPU memory to support vGPU profiles associated with different workloads/vms). Same rationale applies here as above in rejecting claim 6.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Sun et al. U.S. 10,262,390 discloses managing access to a resource pool of GPU under fine grain control.
Ananthanarayanan et al. U.S. 2022/0188569 discloses allocating computing resources during continuous retraining.
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SHIN HON (ERIC) CHEN whose telephone number is (571)272-3789. The examiner can normally be reached Monday to Thursday 9am- 7pm.
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, Lynn Feild can be reached at 571-272-2092. 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.
/SHIN-HON (ERIC) CHEN/Primary Examiner, Art Unit 2431
Prosecution Insights