Skip to content

[TRT RTX EP] Enable double datatype for intermediate nodes#26423

Merged
chilo-ms merged 3 commits into
microsoft:mainfrom
thevishalagarwal:add-double-dtype
Nov 5, 2025
Merged

[TRT RTX EP] Enable double datatype for intermediate nodes#26423
chilo-ms merged 3 commits into
microsoft:mainfrom
thevishalagarwal:add-double-dtype

Conversation

@thevishalagarwal

@thevishalagarwal thevishalagarwal commented Oct 28, 2025

Copy link
Copy Markdown
Contributor
  • Enable double datatype for intermediate nodes
  • add check for unsupported input output nodes in GetCapability()
  • add free dims override for ort perf test in ep context compilation

Comment thread onnxruntime/core/providers/nv_tensorrt_rtx/nv_execution_provider.cc
@thevishalagarwal thevishalagarwal changed the title [TRT RTX EP] Add double datatype support [TRT RTX EP] Add double support for intermediate nodes Oct 29, 2025
@thevishalagarwal thevishalagarwal changed the title [TRT RTX EP] Add double support for intermediate nodes [TRT RTX EP] Enable double datatype for intermediate nodes Oct 29, 2025
@thevishalagarwal

Copy link
Copy Markdown
Contributor Author

@chilo-ms for review

@chilo-ms

chilo-ms commented Nov 5, 2025

Copy link
Copy Markdown
Contributor

/azp run Linux QNN CI Pipeline,Win_TRT_Minimal_CUDA_Test_CI,Windows ARM64 QNN CI Pipeline,Windows GPU Doc Gen CI Pipeline,Windows x64 QNN CI Pipeline

@azure-pipelines

Copy link
Copy Markdown
Azure Pipelines successfully started running 4 pipeline(s).

@chilo-ms
chilo-ms merged commit 82fe431 into microsoft:main Nov 5, 2025
99 of 102 checks passed
Rohanjames1997 pushed a commit to Rohanjames1997/onnxruntime that referenced this pull request Dec 4, 2025
…#26423)

* Enable double datatype for intermediate nodes
* add check for unsupported input output nodes in GetCapability()
* add free dims override for ort perf test in ep context compilation
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