IP Library Granted Patent US 12,524,355
Granted Patent B2
US 12,524,355 · App. 18/134,079 · Granted Jan 13, 2026

Integrated chiplet-based central processing units with accelerators for video processing

Inventors: Harikrishna Madadi Reddy (San Jose, CA); Yunqing Chen (Los Altos, CA); Baheerathan Anandharengan (Milpitas, CA); Christian Markus Petersen (Golden, CO)
Assignee: Meta Platforms, Inc.
G06F13/1663G06F21/44G06F21/572H04N19/436H04N19/85G06F2213/1602G06F2221/033
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,524,355
App. No.
18/134,079
Granted
Jan 13, 2026
Kind
B2
Abstract

In some embodiments, a method includes receiving, at a central processing unit (CPU)-based demultiplexer of a CPU, an input video data stream; performing, at the CPU, an accelerator decoding configuration assessment of an accelerator decoding configuration of an accelerator; and based upon the accelerator decoding configuration assessment, dynamically decoding CPU-based demultiplexer output from the CPU-based demultiplexer using a CPU-based decoding unit and an accelerator-based decoding unit. In some embodiments of the method, the accelerator decoding configuration assessment includes performing an accelerator-based decoding unit hardware configuration assessment of the accelerator-based decoding unit and a CPU-based decoding unit software configuration assessment of the CPU-based decoding unit software utilized for the CPU-based decoding unit.

Claims (41)

1 . A system-on-chip, comprising:

a central processing unit (CPU)-based demultiplexer;

a CPU-based decoding unit and an accelerator-based decoding unit coupled to the CPU-based demultiplexer via a die-to-die interconnect; and

an accelerator-based preprocessing unit coupled to the CPU-based decoding unit and the accelerator-based decoding unit via the die-to-die interconnect, wherein based upon an accelerator decoding configuration assessment, the system-on-chip utilizes the CPU-based decoding unit and the accelerator-based decoding unit to dynamically decode CPU-based demultiplexer output from the CPU-based demultiplexer for video processing;

wherein the accelerator decoding configuration assessment comprises an accelerator-based decoding unit hardware configuration assessment that includes determining whether the CPU-based demultiplexer output maps to an accelerator-based decoding unit hardware configuration of the accelerator-based decoding unit.

2 . The system-on-chip of claim 1 , wherein:

the accelerator decoding configuration assessment further comprises a CPU-based decoding unit software configuration assessment.

3 . The system-on-chip of claim 2 , wherein:

the CPU-based decoding unit software configuration assessment includes determining whether the CPU-based demultiplexer output maps to a CPU-based decoding unit software configuration of the CPU-based decoding unit.

4 . The system-on-chip of claim 3 , wherein:

the system-on-chip is configured to utilize the accelerator-based preprocessing unit to dynamically preprocess decoder output from the CPU-based decoding unit and the accelerator-based decoding unit.

5 . The system-on-chip of claim 4 , wherein:

the system-on-chip utilizes an accelerator-based preprocessing unit configuration assessment to dynamically preprocess decoder output from the CPU-based decoding unit and the accelerator-based decoding unit.

6 . The system-on-chip of claim 5 , wherein:

the accelerator-based processing unit configuration assessment includes an accelerator-based preprocessing unit hardware configuration assessment and a CPU-based preprocessing unit software configuration assessment.

7 . The system-on-chip of claim 6 , further comprising:

an accelerator-based encoding unit coupled to the accelerator-based preprocessing unit, wherein the system-on-chip is configured to utilize the accelerator-based encoding unit to dynamically encode preprocessing output from the accelerator-based preprocessing unit.

8 . The system-on-chip of claim 7 , wherein:

the system-on-chip utilizes an accelerator-based encoding unit hardware configuration assessment to dynamically encode preprocessor output from the accelerator-based preprocessing unit.

9 . A system-on-chip, comprising:

a central processing unit (CPU)-based demultiplexer;

a CPU-based decoding unit and an accelerator-based decoding unit coupled to the CPU based demultiplexer via a die-to-die interconnect; and

an accelerator-based preprocessing unit coupled to the CPU-based decoding unit and the accelerator-based decoding unit via the die-to-die interconnect, wherein based upon an accelerator decoding configuration assessment, the system-on-chip utilizes the CPU-based decoding unit and the accelerator-based decoding unit to dynamically decode CPU-based demultiplexer output from the CPU-based demultiplexer for video processing;

wherein the accelerator decoding configuration assessment comprises a CPU-based decoding unit software configuration assessment that includes determining whether the CPU-based demultiplexer output maps to a CPU-based decoding unit software configuration of the CPU-based decoding unit.

10 . The system-on-chip of claim 9 , wherein the accelerator decoding configuration assessment further comprises an accelerator-based decoding unit hardware configuration assessment.

11 . The system-on-chip of claim 10 , wherein the accelerator-based decoding unit hardware configuration assessment that includes determining whether the CPU-based demultiplexer output maps to an accelerator-based decoding unit hardware configuration of the accelerator-based decoding unit.

12 . The system-on-chip of claim 9 , wherein the system-on-chip is configured to utilize the accelerator-based preprocessing unit to dynamically preprocess decoder output from the CPU-based decoding unit and the accelerator-based decoding unit.

13 . The system-on-chip of claim 12 , wherein the system-on-chip utilizes an accelerator-based preprocessing unit configuration assessment to dynamically preprocess decoder output from the CPU-based decoding unit and the accelerator-based decoding unit.

14 . The system-on-chip of claim 13 , wherein the accelerator-based processing unit configuration assessment includes an accelerator based preprocessing unit hardware configuration assessment and a CPU-based preprocessing unit software configuration assessment.

15 . The system-on-chip of claim 14 , further comprising:

an accelerator-based encoding unit coupled to the accelerator-based preprocessing unit, wherein the system-on-chip is configured to utilize the accelerator-based encoding unit to dynamically encode preprocessing output from the accelerator based preprocessing unit.

16 . The system-on-chip of claim 15 , wherein:

the system-on-chip utilizes an accelerator-based encoding unit hardware configuration assessment to dynamically encode preprocessor output from the accelerator-based preprocessing unit.

17 . A system-on-chip, comprising:

a central processing unit (CPU)-based demultiplexer;

a CPU-based decoding unit and an accelerator-based decoding unit coupled to the CPU based demultiplexer via a die-to-die interconnect; and

an accelerator-based preprocessing unit coupled to the CPU-based decoding unit and the accelerator-based decoding unit via the die-to-die interconnect, wherein based upon an accelerator decoding configuration assessment, the system-on-chip utilizes the CPU-based decoding unit and the accelerator-based decoding unit to dynamically decode CPU-based demultiplexer output from the CPU-based demultiplexer for video processing;

wherein the system-on-chip is configured to utilize the accelerator-based preprocessing unit to dynamically preprocess decoder output from the CPU-based decoding unit and the accelerator-based decoding unit.

18 . The system-on-chip of claim 17 , wherein the accelerator decoding configuration assessment further comprises an accelerator-based decoding unit hardware configuration assessment and a CPU-based decoding unit software configuration assessment.

19 . The system-on-chip of claim 18 , wherein the CPU-based decoding unit software configuration assessment includes determining whether the CPU-based demultiplexer output maps to a CPU-based decoding unit software configuration of the CPU-based decoding unit.

20 . The system-on-chip of claim 17 , wherein the accelerator-based processing unit configuration assessment includes an accelerator-based preprocessing unit hardware configuration assessment and a CPU-based preprocessing unit software configuration assessment.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Aug 9, 2023
From: REDDY, HARIKRISHNA MADADI; CHEN, YUNQING; ANANDHARENGAN, BAHEERATHAN; PETERSEN, CHRISTIAN MARKUS
To: META PLATFORMS, INC.
Reel/Frame 064542/0769 →
Continuity (2)
Provisional Application 63436543 · Dec 31, 2022
Related Publication 20240223789A1 · Jul 4, 2024
References Cited (33)
US 10694189B1 · Seshadri et al. · 2020 [cited by applicant]
US 11030117B2 · Jayasena et al. · 2021 [cited by applicant]
US 11263143B2 · Gupta et al. · 2022 [cited by applicant]
US 11315328B2 · Aksoy et al. · 2022 [cited by applicant]
US 11451790B2 · Liu et al. · 2022 [cited by applicant]
US 20090060032A1 · Schmit et al. · 2009 [cited by applicant]
US 20120320967A1 · Gao · 2012 [cited by examiner]
US 20130156101A1 · Lu et al. · 2013 [cited by applicant]
US 20130215978A1 · Wu · 2013 [cited by examiner]
US 20150237356A1 · Wu et al. · 2015 [cited by applicant]
US 20180103261A1 · Sun · 2018 [cited by examiner]
US 20180183573A1 · Ghosh et al. · 2018 [cited by applicant]
US 20190042425A1 · Shifer · 2019 [cited by applicant]
US 20190236022A1 · Gopal et al. · 2019 [cited by applicant]
US 20200059435A1 · Yang · 2020 [cited by applicant]
US 20210096840A1 · Kotary et al. · 2021 [cited by applicant]
US 20210294772A1 · Srivastava et al. · 2021 [cited by applicant]
US 20220197506A1 · Blagodurov et al. · 2022 [cited by applicant]
US 20220222194A1 · Chandwani et al. · 2022 [cited by applicant]
US 20220245252A1 · Smith et al. · 2022 [cited by applicant]
US 20220334972A1 · Vijayrao et al. · 2022 [cited by applicant]
US 20220334995A1 · Das Sharma et al. · 2022 [cited by applicant]
US 20220358208A1 · Vijayrao et al. · 2022 [cited by applicant]
US 20240103827A1 · Bisa et al. · 2024 [cited by applicant]
EP 3138040A1 · 2017 [cited by applicant]
WO 2017062237A1 · 2017 [cited by applicant]
WO 2022133053A1 · 2022 [cited by applicant]
International Preliminary Report on Patentability for International Application No. PCT/US2023/083763, mailed Jul. 10, 2025, 9 pages. [cited by applicant]
International Preliminary Report on Patentability for International Application No. PCT/US2023/084599, mailed Jul. 10, 2025, 6 pages. [cited by applicant]
International Search Report and Written Opinion for International Application No. PCT/US2023/083763, mailed Apr. 4, 2024, 12 pages. [cited by applicant]
International Search Report and Written Opinion for International Application No. PCT/US2023/084599, mailed Apr. 17, 2024, 9 pages. [cited by applicant]
International Preliminary Report on Patentability for International Application No. PCT/US2023/084685, mailed Jul. 10, 2025, 7 pages. [cited by applicant]
International Search Report and Written Opinion for International Application No. PCT/US2023/084685, mailed Apr. 18, 2024, 10 pages. [cited by applicant]