Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Optimize GPU decoder #4351

Merged
merged 1 commit into from
Nov 7, 2022
Merged

Optimize GPU decoder #4351

merged 1 commit into from
Nov 7, 2022

Conversation

JanuszL
Copy link
Contributor

@JanuszL JanuszL commented Oct 13, 2022

  • optimizes GPU decoder
  • introduces reusing of the NVDEC instances across the iterations
    and inside the batch
  • refactors the NVDEC decoder instance creation to pull
    parameters from the parser
  • adds RAII pattern for the parser handle

Signed-off-by: Janusz Lisiecki [email protected]

Category:

Other (e.g. Documentation, Tests, Configuration)

Description:

  • optimizes in memory GPU decoder
  • introduces reusing of the NVDEC instances across the iterations
    and inside the batch
  • refactors the NVDEC decoder instance creation to pull
    parameters from the parser
  • adds RAII pattern for the parser handle

Additional information:

Affected modules and functionalities:

Key points relevant for the review:

  • NA

Tests:

  • Existing tests apply
    • CfrFrameRateMpeg4NoIndex
    • VfrFrameRateMpeg4NoIndex
    • test_video_decoder
  • New tests added
    • Python tests
    • GTests
    • Benchmark
    • Other
  • N/A

Checklist

Documentation

  • [] Existing documentation applies
  • Documentation updated
    • Docstring
    • Doxygen
    • RST
    • Jupyter
    • Other
  • N/A

DALI team only

Requirements

  • Implements new requirements
  • Affects existing requirements
  • N/A

REQ IDs: N/A

JIRA TASK: N/A

@JanuszL JanuszL marked this pull request as draft October 13, 2022 17:44
@JanuszL JanuszL force-pushed the decoder_gpu_opt branch 4 times, most recently from 98696fd to 9ec434b Compare October 18, 2022 05:45
@JanuszL JanuszL changed the title Decoder gpu opt Optimize in memory GPU decoder Oct 25, 2022
@JanuszL JanuszL changed the title Optimize in memory GPU decoder Optimize memory GPU decoder Oct 25, 2022
@JanuszL JanuszL marked this pull request as ready for review October 26, 2022 08:50
@awolant awolant self-assigned this Oct 26, 2022
@JanuszL JanuszL changed the title Optimize memory GPU decoder Optimize GPU decoder Oct 26, 2022
@JanuszL
Copy link
Contributor Author

JanuszL commented Oct 27, 2022

!build

@dali-automaton
Copy link
Collaborator

CI MESSAGE: [6310323]: BUILD STARTED

@dali-automaton
Copy link
Collaborator

CI MESSAGE: [6310323]: BUILD FAILED

@JanuszL JanuszL force-pushed the decoder_gpu_opt branch 2 times, most recently from b7e576e to 516c0e4 Compare October 28, 2022 21:03
@JanuszL
Copy link
Contributor Author

JanuszL commented Oct 28, 2022

!build

@dali-automaton
Copy link
Collaborator

CI MESSAGE: [6324018]: BUILD STARTED

@dali-automaton
Copy link
Collaborator

CI MESSAGE: [6324018]: BUILD FAILED

@dali-automaton
Copy link
Collaborator

CI MESSAGE: [6324018]: BUILD PASSED

- optimizes GPU decoder
- introduces reusing of the NVDEC instances across the iterations
  and inside the batch
- refactors the NVDEC decoder instance creation to pull
  parameters from the parser
- adds RAII pattern for the parser handle

Signed-off-by: Janusz Lisiecki <[email protected]>
@JanuszL
Copy link
Contributor Author

JanuszL commented Nov 7, 2022

!build

@dali-automaton
Copy link
Collaborator

CI MESSAGE: [6411226]: BUILD STARTED

@dali-automaton
Copy link
Collaborator

CI MESSAGE: [6411226]: BUILD FAILED

@dali-automaton
Copy link
Collaborator

CI MESSAGE: [6411226]: BUILD PASSED

@JanuszL JanuszL merged commit c9eb44e into NVIDIA:main Nov 7, 2022
@JanuszL JanuszL deleted the decoder_gpu_opt branch November 7, 2022 22:37
@JanuszL JanuszL mentioned this pull request Jan 11, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants