Detailed Action
This action is in response to amendments filed on 12/19/2025.
This application was filed on 05/30/2023.
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claims 1-6, 8-13, 15-20 are pending.
Claims 1-6, 8-13, 15-20 are rejected.
Applicant's Response
In Applicant's Response dated 12/19/2025, Applicant amended claims 1-6, 8-13, 15-20, canceled claims 7, 14, and 21, and amended specification. Applicant argued against various rejections previously set forth in the Office Action mailed on 09/22/2025.
In light of Applicant' s amendments and remarks, all objections to the specification are withdrawn.
In light of Applicant' s amendments and remarks, all rejections of claims under 35 U.S.C. 101 set forth previously are withdrawn.
Examiner Notes
Examiner cites particular columns, paragraphs, figures and line numbers in the references as applied to the claims below for the convenience of the applicant. Although the specified citations are representative of the teachings in the art and are applied to the specific limitations within the individual claim, other passages and figures may apply as well. It is respectfully requested that, in preparing responses, the applicant fully consider the references in their entirety as potentially teaching all or part of the claimed invention, as well as the context of the passage as taught by the prior art or disclosed by the examiner.
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 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.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim 1-2, 4-6, 8-9, 11-13, 15-16, 18-20 are rejected under 35 U.S.C. 103 as being unpatentable over Fontoura et al. (US 20190163517 A1, referred hereinafter as D1) in view of Baggerman (US 20190139185 A1, referred hereinafter as D2) in view of Kaufer et al. (US 20170192809 A1, referred hereinafter as D4).
As per claim 1, D1 discloses,
A method comprising, (D1, abstract).
collecting, by a right-sizing engine running within [a system], graphics processing unit (GPU) usage data pertaining to usage of a GPU by the VM, (D1, 0029, 0033, 0035 and figure 2D shows/discloses collecting, by a right-sizing engine/prediction model/controller running within the system of D1, data (e.g. customer, tenant, allocation, usage data) pertaining to usage of a graphics processing unit (GPU) (e.g. allocated resource includes GPUs and memory) on which the VM is placed on/executed).
wherein the collected GPU usage data includes an amount GPU compute or GPU memory resources consumed by the VM over a previous period, (D1, 0029, 0033, 0035 and figure 2D shows/discloses collecting/receiving historical/previous period resource/GPU/memory usage for data for one or more VMs for learning purposes to train predictive engine.).
analyzing, by the right-sizing engine, the collected GPU usage data to predict a maximum amount of GPU resources for executing… the VM during a future period, wherein the maximum amount GPU resources includes a maximum amount of GPU computer or GPU memory resources to consumed by the VM, (D1, 0029, 0033, 0035, 0045 and figure 2D shows/discloses collecting, by a right-sizing engine/prediction model running within the system of D1, data (e.g. customer, tenant, allocation, usage data) to learn/generate prediction model, and based collection data/inputs into the prediction model, the prediction model calculates/predicts a maximum amount of resources including GPU resources that the VM will need to execute and execute VM in future period according to generated configuration based on the prediction.).
determining, by the right-sizing engine, a right-sized GPU profile for the VM based on the predicted maximum amount of GPU resources, (D1, 0029, 0033- 0035, 0040 and figure 2D shows/discloses calculating/predicting a maximum amount of resources including GPU resources that the VM will likely require during its runtime and accordingly, determining/generating, by the right-sizing engine(e.g. model/rightsizing controller), a right-sized GPU profile (e.g. configuration, per step 220 of figure 2D) for the VM based on the predicted maximum amount of GPU resources).
wherein the determined GPU specifies an amount GPU processing cores or GPU memory to be reserved for the VM for executing … during the further period, (D1, 0027, 0029, 0033- 0035, 0040 and figure 2D shows/discloses deployment configuration specifies GPU/memory allocation to executing the VM, and calculating/predicting a maximum amount of resources including GPU resources and/or memory that the VM will likely require during its runtime and accordingly, determining/generating, by the right-sizing engine(e.g. model/rightsizing controller), a right-sized GPU profile (e.g. modified/new configuration, per step 220 of figure 2D) for the VM based on the predicted maximum amount of GPU resources/memory).
resizing the VM to cause the VM to execute… on the amount of GPU processing cores or GPU memory specified by the determined GPU profile, [deploying] the VM with determined GPU profile, (D1, 0029, 0033, 0035, 0045 and figure 2D shows/discloses collecting, by a right-sizing engine/prediction model running within the system of D1, data (e.g. customer, tenant, allocation, usage data) to learn/generate prediction model, and based collection data/inputs into the prediction model, the prediction model calculates/predicts a maximum amount of resources including GPU resources that the VM will need to execute and execute VM workloads in future period according to generated configuration based on the prediction, where as shown in figure 2D, D1 resizing/executing/deploying the VM to cause the VM to execute the one or more workloads on the amount of GPU processing cores or GPU memory specified by the determined GPU profile/configuration.).
D1 discloses a right-sizing engine running within system of D1; however, D1 fails to expressly disclose – [engine/application] running within a virtual machine (VM), and one or more workloads [of VM].
However, the design choice to have engine/application running within a virtual machine (VM) was notoriously well known. For instance, D2 (0033) discloses engine/applications/workloads running within a virtual machine (VM).
Accordingly, it would have been obvious to one having ordinary skill in the art before the effective filing date of the invention, disclosed in D1, to include engines/application and/or workloads running within a virtual machine (VM). This would have been obvious with predicable results of running an engine/application within a VM as known in the art and disclosed by D2.
D1/D2 discloses workloads/application; however, D1 fails to expressly disclose - and wherein resizing the VM comprises saving a runtime states of the one or more workloads, restarting the VM with… , and restoring the saved runtime state for each of the one or more workloads in the restarted VM for execution by the restarted VM.
However, D4 (0036) discloses known methods of having a disk image and/or memory state of VM/workloads saved such that the VM/workload can easily be restored without requiring a blank-state VM which reads on wherein resizing the VM comprises saving a runtime states of the one or more workloads, restarting the VM with… , and restoring the saved runtime state for each of the one or more workloads in the restarted VM for execution by the restarted VM.
Accordingly, it would have been obvious to one having ordinary skill in the art before the effective filing date of the invention, disclosed in D1, to include the teachings of D4 as noted above. This would have been obvious with predicable results of easily having VM/workloads restored without requiring a blank-state VM as disclosed by D4.
As per claim 2, the rejection of claim 1 further incorporated, D1 discloses
wherein the VM shares use of the GPU with other VMs, (D1, 0003, 0029-0030 the VM shares use of the GPU/resources with other VMs.).
and the analyzing includes predicating a maximum amount of [memory] to be consumed by the VM, (D1, 0029, 0033, 0035 and figure 2D shows/discloses collecting, by a right-sizing engine/prediction model running within the system of D1, data (e.g. customer, tenant, allocation, usage data) to learn/generate prediction model, and based collection data/inputs into the prediction model, the prediction model calculates/predicts a maximum amount of resources including a maximum amount of video RAM/resource that the VM will likely require during its runtime.).
D1 arguably/inherently discloses VMs using virtual GPU sharing; nevertheless, for sake of completeness, D2 (0010) discloses - VMs using virtual GPU sharing in that D2 discloses a vGPU profile indicates how virtualized resources (e.g. virtual GPU) of a physical GPU may be allocated to/shared with virtual machines supported by a node in which the physical GPU is located.
Accordingly, it would have been obvious to one having ordinary skill in the art before the effective filing date of the invention, disclosed in D1, to include VMs using virtual GPU sharing. This would have been obvious with predicable results of allocating various VMs to specific vGPU profiles/virtual resources/virtual GPU to achieve greater allocation efficiency as disclosed by D2(0010).
D1 fails to expressly disclose - video random access memory (VRAM).
However, the examiner takes official notice that video random access memory (VRAM) was notoriously well known before effective filing of the invention.
Accordingly, it would have been obvious to one having ordinary skill in the art before the effective filing date of the invention, disclosed in D1, to include video random access memory (VRAM). This would have been obvious with predicable results of additionally using video random access memory (VRAM) for storing data/state as known in the art.
As per claim 4, the rejection of claim 1 further incorporated, D1 discloses,
wherein the analyzing comprises: fitting the collected GPU usage data to distribution, (D1, 0029, 0033, 0035-0037 and figure 2B, 2D shows/discloses collecting, by a right-sizing engine/prediction model running within the system of D1, data (e.g. customer, tenant, allocation, usage data) to learn/generate prediction model, and where analysis includes fitting the collected data to a data distribution/bucketizing).
and identifying one or more data values located a or above a threshold percentile of the data distribution, as the precited maximum amount of GPU resources, (D1, 0029, 0033, 0035-0037 and figure 2D shows/discloses collecting, by a right-sizing engine/prediction model running within the system of D1, data (e.g. customer, tenant, allocation, usage data) to learn/generate prediction model, and where analysis includes fitting the collected data to a data distribution/bucketizing and grouping/predicting the maximum amount of resources which includes GPU resources based on data values located above a threshold percentile of the data distribution as the predicted maximum amount of GPU (e.g. prediction of VM resource requirement is based on selected ranges including upper portion of the distribution as the maximum)).
As per claim 5, the rejection of claim 1 further incorporated, D1 discloses,
wherein the analyzing comprises: training a machine learning (ML) model based on the collected data GPU usage data; and providing at least a portion of the collected data GPU usage data as input to the trained ML model, the model predicting the maximum amount of GPU resources based on at least a portion of the collected GPU usage data input thereto, (D1, 0029, 0033, 0035 and figure 2D shows/discloses collecting, by a right-sizing engine/prediction model running within the system of D1, data (e.g. customer, tenant, allocation, usage data) to train/learn/generate prediction model, and based collection data/inputs into the prediction model, the prediction model calculates/predicts a maximum amount of resources including GPU resources that the VM will likely require during its runtime).
As pe claim 6, the rejection of claim 1 further incorporated, D1 discloses,
saving the determined GPU profile in manner that enables later retrieval and presentation to a creator of the VM, (D1, 0006, 0033-0035, 0040, 0078 and figure 2D shows/discloses generating/saving deployment configuration (e.g. right-sized GPU profile) and communicate to the interface of user/creator/deployer to VM, the deployment configuration.)
As per claims 8-9, 11-13, 15-16, 18-20: Claims 8-9, 11-13, 15-16, 18-20 are medium and system claims corresponding to method claims 1-2, 4-6, and are of substantially same scope.
Accordingly, claims 8-9, 11-13, 15-16, 18-20 are rejected under the same rational as set forth for claims 1-2, and 4-6.
Claim 3, 10, and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Fontoura et al. (US 20190163517 A1, referred hereinafter as D1) in view of Baggerman (US 20190139185 A1, referred hereinafter as D2) in view of Kaufer et al. (US 20170192809 A1, referred hereinafter as D4) in view Duluk et al. (US 20230288471 A1, referred hereinafter as D3).
As per claim 3, the rejection of claim 1 further incorporated, D1 discloses,
wherein the VM shares use of the GPU with other VMs, (D1, 0003, 0029-0030 the VM shares use of the GPU/resources with other VMs.).
and the analyzing includes predicating a maximum amounts of GPU computer and GPU memory resources to be consumed, (D1, 0029, 0033, 0035 and figure 2D shows/discloses collecting, by a right-sizing engine/prediction model running within the system of D1, data (e.g. customer, tenant, allocation, usage data) to learn/generate prediction model, and based collection data/inputs into the prediction model, the prediction model calculates/predicts a maximum amount of resources including wherein the analyzing predicts a maximum amount of GPU memory/RAM resources and a maximum amount of GPU compute resources that the VM will likely require during its runtime.).
D1 fails to expressly disclose - using multi-instance GPU (MIG).
D3(0031) discloses using multi-instance GPU (MIG).
Accordingly, it would have been obvious to one having ordinary skill in the art before the effective filing date of the invention, disclosed in D1, to include using multi-instance GPU (MIG). This would have been obvious with predicable results of allowing the GPU to be securely partitioned into many separate GPU Instances and providing multiple users with separate GPU resources to accelerate their respective applications/VMs as disclosed by D3 (0031).
As per claims 10, and 17: Claims 10, and 17 are medium/system claims corresponding to method claim 3, and is of substantially same scope.
Accordingly, claims 10, and 17 is rejected under the same rational as set forth for claim 3.
Response to Arguments
Applicant’s arguments filed on 12/19/2025 have been fully considered but they are not persuasive and/or moot in view of new/modified grounds of rejections.
Conclusion
Applicant's amendment necessitated any new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). 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 extension fee 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 date of this final action.
The prior art made of record and not relied upon is considered pertinent to applicant’s disclosure.
METHOD AND SYSTEM THAT ALLOCATES VIRTUAL NETWORK COST IN A SOFTWARE-DEFINED DATA CENTER
DOCUMENT ID
US 20160203528 A1
DATE PUBLISHED
2016-07-14
Abstract
The present disclosure describes methods and systems that allocate costs of deploying and operating a virtual network to tenants that use the virtual network. In one implementation, costs are allocated to tenant virtual machines (“VMs”) by determining a network bandwidth of a virtual network, determining a common cost of operating the virtual network, determining a service capacity for each network service provided by the virtual network, and determining a service cost for each network service. A portion of the common cost is allocated to each VM based on the proportion of network bandwidth used by each VM, and a portion of the service cost is allocated to each VM based on the proportion of the service capacity used by each VM.
See form 892.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MUSTAFA A AMIN whose telephone number is (571)270-3181. The examiner can normally be reached on Monday-Friday from 8:00 AM to 5:00 PM.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Kevin Young, can be reached on 571-270-3180. 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 Patent Center. Status information for published applications may be obtained from Patent Center. Status information for unpublished applications is available through Patent Center for authorized users only. Should you have questions about access to Patent Center, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free).
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) Form at https://www.uspto.gov/patents/uspto-automated- interview-request-air-form.
/MUSTAFA A AMIN/ Primary Examiner, Art Unit 2194