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 .
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.
Drawings
The replacement drawings for figure 2 were received on 02/05/2026. These drawings are accepted.
Response to Arguments
This is in response to applicant’s amendment/response filed on 20/05/2026 which have been entered and made of record.
Applicant’s arguments regarding the objection to the drawings have been fully considered and are persuasive. The objection to the drawings has been overcome, the objection has been withdrawn.
Applicant’s arguments regarding claim rejections under 35 U.S.C. 103 have been fully considered but they are not persuasive.
Applicant argues while the Office at page of the Office Action concedes that the '951 publication does not explicitly show a "predict token," the Office nonetheless asserts that the '951 publication "teaches utilizing frame data or commands that indicate if a fame is extrapolated or interpolated." Assignee respectfully disagrees. The '951 publication states that it shows enabling time-based predication with frame generation and scaling via artificial intelligence (see, for example, paragraph [0002] of the '951 publication). The '951 continues to state that:
When predication for queued command buffers is enabled, command buffers for workloads that are not related to the workload that is generated in response to the user input are bypassed. High priority command buffers that include workloads generated in response to user input can then be executed immediately, without having to wait for previously enqueued workloads to complete. ['951 publication, para. 0058, emphasis added]
The '951 publication further states that it shows that a target frame pacing using neural frame generation and time-based predication means that for a frame a determination can be made based on the difference between the time in which the command buffer for frame would begin processing and the deadline for the next frame update. At this time, the commands within the command buffer for that frame can be bypassed via predication. Instead of processing the commands for the bypassed frame, a frame generation request can be submitted to perform neural frame generation based on extrapolation. [see, e.g., '951 publication, paragraph 0405] The '951 publication further states that it shows that:
time-based predication enables the commands of a command buffer to be skipped based on a determination that the commands for a command buffer that is ready to execute are not likely to complete before some predetermined time, for example, in time for display at the next display update interval. In place of those commands, an Al draw command can be executed, which will have a shorter and more deterministic execution time. ['951 publication, paragraph 0412]
As such, it is submitted that '951 appears to discuss describes bypassing, or skipping, frames that will not be completed on time and implementing a neural frame generation via an Al draw command to generate the bypassed, or skipped, frame using interpolation or extrapolation. As such, it is submitted that the '951 publication does not disclose or suggest that a predict token "is indicative of an extrapolated and/or interpolated target frame to be generated for the sequence of frames; [and] ... generating the extrapolated and/or interpolated target frame for the sequence of frames based on the execution of the predict token" as set forth in claim 1. As such, the '951 publication does not impart the benefit of enabling a GPU to generate, or render, a higher number of frames for the application executing on the CPU, thereby reducing latency and/or increasing the frame rate. Accordingly, it is submitted that claim 1 patentably distinguishes over claim 1.
Examiner respectable disagrees. According to BRI (Broadest Reasonable Interpretation) in the context of a GPU a token can be a unit of data processed by a model, cryptocurrency, a command, etc... In claim 1 a “predict token” is received from a command stream and indicates an extrapolated and/or interpolated target frame to be generated for a sequence of frames. Examiner is interpreting a “predict token” to be either a unit of data (frame data) predicted by a model or a predict command (command). Regarding the reference Panneer (‘951), Panneer teaches predicting interpolated or extrapolated frames from frame data, optical flow data, and timestamps.(Para. 0395) to reduce latency by maintaining a consistent frame pacing (Para. 0394). Panneer determines whether to interpolate or extrapolate a frame based on the results of a trained model and a users’ input(Para. 0395 and 0404) which is a command. Thus, Panneer fails to explicitly teach the terminology of “predict token”. However Panneer teaches commands and frame data that function the same as the “predict token” claimed in claim 1.
Regarding the remaining arguments applicant argues with respect to the amended claim language, which is fully addressed in the prior art rejections set forth below.
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(s) 1-20 are rejected under 35 U.S.C. 103 as being unpatentable over Panneer et al. U.S. Patent Application Publication US 20240311951 A1 (hereinafter Panneer).
Regarding claim 1, Panneer teaches a method of operating a graphics processing unit that renders a sequence of frames, the method comprising: (Para. 0058 and 0073)
receiving a predict token (Frame Data or Commands) via a rendering command stream (Para. 0285) for the graphics processing unit (Para. 0058 and 0073), wherein the predict token (Frame Data or Commands) is indicative of an extrapolated and/or interpolated target frame to be generated for the sequence of frames; (Para. 0394-0395 and 0403-0404, Fig. 30A and Fig. 30B)
executing the predict token (Frame Data or Commands); (Para. 0396-0401)
generating the extrapolated and/or interpolated target frame for the sequence of frames based on the execution of the predict token; (Para. 0403-0406)
and storing the generated extrapolated and/or interpolated target frame in a frame buffer (Buffer Queue or Frame Buffer, Para. 0075, 0262, and 0440).
Panneer fails to explicitly teach a predict token.
However, Paneer teaches utilizing frame data or commands that indicate if a frame is extrapolated or interpolated (Para. 0394-0395 and 0403-0404, Fig. 30A and Fig. 30B).
As stated above, according to BRI (Broadest Reasonable Interpretation) in the context of a GPU a token can be a unit of data processed by a model, cryptocurrency, a command, etc... In claim 1 a “predict token” is received from a command stream and indicates an extrapolated and/or interpolated target frame to be generated for a sequence of frames. Examiner is interpreting a “predict token” to be either a unit of data (frame data) predicted by a model or a predict command (command). Regarding the reference Panneer (‘951), Panneer teaches predicting interpolated or extrapolated frames from frame data, optical flow data, and timestamps.(Para. 0395) to reduce latency by maintaining a consistent frame pacing (Para. 0394). Panneer determines whether to interpolate or extrapolate a frame based on the results of a trained model and a users’ input(Para. 0395 and 0404) which is a command. Thus, Panneer fails to explicitly teach the terminology of “predict token”. However Panneer teaches commands and frame data that function the same as the “predict token” claimed in claim 1.
Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Panneer’s GPU to incorporate a predict token. Since doing so would provide the benefit of separating out certain frame data into tokens to increase the efficiency of the GPU.
Regarding claim 2, Panneer teaches the method of claim 1, in which the predict token references:
a predict command buffer (Para. 0262), wherein the predict command buffer includes one or more predict commands (Para. 0404, 0406, and 0416) defining one or more operations (How Data is Processed, Para. 0073) relating to the extrapolated and/or interpolated target frame to be generated by the graphics processing unit; (Para. 0058 and 0404)
and a predict parameter buffer (Parameters), wherein the predict parameter buffer includes one or more parameters that store data relating to the extrapolated and/or interpolated target frame to be generated by the graphics processing unit. (Para. 0073 and 0341-0342)
Panneer fails to explicitly teach a predict parameter buffer.
Panneer teaches the command buffer includes the parameters that relate to the frame to be generated (Para. 0072-0073 and 0341-0342). As well as their GPU utilizes multiple different buffers to store information (Para. 0073 and 0075). Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Panneer’s buffers to incorporate a buffer for parameters. Since doing so would provide the benefit of storing parameters alone instead of together with the commands.
Regarding claim 3, Panneer teaches the method of claim 2, in which executing, by the graphics processing unit, the predict token, further comprises:
executing the one or more predict commands (Para. 0404, 0406, and 0416) of the predict command buffer (Para. 0262) and populating one or more parameters (Para. 0073 and 0341-0342) of the predict parameter buffer based on one or more results of the execution of the one or more predict commands.
Regarding claim 4, Panneer teaches the method of claim 2, in which the one or more operations include one or more of Motion Vector (Para. 0432) Estimation MVE, Asynchronous Reprojection ARP, Asynchronous Time Warp ATW, Asynchronous Space Warp ASW, Lens Matched Shading LMS, Variable Rate Shading VRS, Fragment Density Map FDM, Foveated Rendering, and Optical Flow Estimation (Para. 0394-0395, 0432, and 0453).
Regarding claim 5, Panneer teaches the method of claim 1, further comprising:
identifying, by the graphics processor unit (Para. 0058 and 0073), a predict frame generation type of the extrapolated and/or interpolated target frame to be generated, wherein the predict frame generation type includes extrapolation and/or interpolation; (Para. 0395 and 0404, Fig. 30A or Fig. 30B)
and generating, by the graphics processor unit (Para. 0058 and 0073), the extrapolated and/or interpolated target frame according to the identified predict frame generation type. (Para. 0395 and 0404, Fig. 30A or Fig. 30B)
Regarding claim 6, Panneer teaches the method of claim 5, in which if the predict frame generation type is identified as extrapolation, the extrapolated and/or interpolated target frame is generated based on a first frame of the sequence of frames and a second frame of the sequence of frames, wherein the first frame and the second frame precede the extrapolated and/or interpolated target frame.(Para. 0395-0397, 0405, and Fig.30A)
Regarding claim 7, Panneer teaches the method of claim 6, in which if the predict frame generation type is identified as interpolation, the extrapolated and/or interpolated target frame is generated based on a first frame of the sequence of frames and a second frame of the sequence of frames, wherein the first frame precedes the extrapolated and/or interpolated target frame and the second frame is subsequent to the extrapolated and/or interpolated target frame. (Para. 0395, 0398-0399, 0406, and Fig. 30B)
Regarding claim 8, Panneer teaches the method of claim 2, in which the one or more parameters (Para. 0073 and 0341-0342) of the predict parameter buffer includes one or more of a motion vector (Para. 0432), a translation (Para. 0084), a position difference, a rotation (Para. 0207), a zoom in/out, a reprojection, a shift, and a 3D estimation. (Geometry Data, Para. 0285)
Regarding claim 9, has similar limitations as of claim 1, therefore it is rejected under the same rationale as claim 1.
Regarding claim 10, has similar limitations as of claim 2, therefore it is rejected under the same rationale as claim 2.
Regarding claim 11, has similar limitations as of claim 3, therefore it is rejected under the same rationale as claim 3.
Regarding claim 12, has similar limitations as of claim 4, therefore it is rejected under the same rationale as claim 4.
Regarding claim 13, has similar limitations as of claim 5, therefore it is rejected under the same rationale as claim 5.
Regarding claim 14, has similar limitations as of claim 6, therefore it is rejected under the same rationale as claim 6.
Regarding claim 15, has similar limitations as of claim 7, therefore it is rejected under the same rationale as claim 7.
Regarding claim 16, has similar limitations as of claim 8, therefore it is rejected under the same rationale as claim 8.
Regarding claim 17, Panneer teaches the graphics processing unit of claim 9, further comprising:
a command stream frontend, wherein the command stream frontend includes a hardware interface, the command stream frontend is operable to: (Para. 0068, 0073, and 0330)
receive an instruction relating to an execution of the predict token(Frame Data or Commands); (Para. 0394-0395)
and execute in hardware the received instruction to generate the extrapolated and/or interpolated target frame based on the predict token(Frame Data or Commands). (Para. 0396-0404)
Regarding claim 18, Panneer teaches the graphics processing unit of claim 9, further comprising:
a command stream frontend, wherein the command stream frontend includes a micro-controller unit (Para. 0397), the command stream frontend is operable to: (Para. 0068, 0073, and 0330)
receive an instruction relating to an execution of the predict token(Frame Data or Commands); (Para. 0073, 0330, and 0394-0395)
and execute in software (Para. 0408) the received instruction to generate the extrapolated and/or interpolated target frame based on the predict token (Frame Data or Commands). (Para. 0396-0404)
Regarding claim 19, Panneer teaches a method of generating a predict token (Frame Data or Commands), wherein the predict token (Frame Data or Commands) is indicative of an extrapolated and/or interpolated target frame to be generated for a sequence of frames, comprising: (Para. 0394-0395)
generating a predict command buffer (Para. 0262), wherein the predict command buffer include one or more commands(Para. 0404, 0406, and 0416) for generating a target frame for the sequence of frames; (Para. 0058 and 0074)
generating a predict parameter buffer (Parameters), wherein the predict parameter buffer includes one or more parameters for generating a target frame for the sequence of frames; (Para. 0072-0073 and 0341-0342)
including one or more instructions that include a reference to a memory location (Para. 0270, 0272, and 0285) of the generated predict command buffer (Para. 0262) and the generated predict parameter buffer in the predict token (Frame Data or Commands); (Para. 0404, 0406, and 0416)
and submitting the generated predict token (Frame Data or Commands) to a rendering command stream (Para. 0285) for a graphics processing unit. (Para. 0403-0404)
Panneer fails to explicitly teach a predict token and a predict parameter buffer.
Panneer teaches the command buffer includes the parameters that relate to the frame to be generated (Para. 0072-0073 and 0341-0342). The GPU utilizes multiple different buffers to store information (Para. 0073, 0075). As well as utilizing frame data or commands that indicate if a frame is extrapolated or interpolated (Para. 0394-0395 and 0403-0404, Fig. 30A and Fig. 30B). Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Panneer’s GPU to incorporate a predict token and a predict parameter buffer. Since doing so would provide the benefit of separating out certain frame data into tokens to increase the efficiency of the GPU and storing parameter information alone instead of together with the commands.
Regarding claim 20, Panneer teaches the method of claim 19, further comprising:
determining one or more performance requirements; (Para. 0245 and 0461)
monitoring the performance requirements; (Para. 0409 and 0440)
and generating the predict token (Frame Data or Commands, Para. 0409 and 0440) based on the monitored performance requirements. (Para. 0289 and 0377)
Conclusion
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 BRIANNA R COCHRAN whose telephone number is (571)272-4671. The examiner can normally be reached Mon-Fri. 7:30am - 5:00pm.
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, Alicia Harrington can be reached at (571) 272-2330. 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.
/BRIANNA RENAE COCHRAN/Examiner, Art Unit 2615
/ALICIA M HARRINGTON/Supervisory Patent Examiner, Art Unit 2615
Prosecution Insights