Limit the Pipeline ability to build cuda 11

.github/ISSUE_TEMPLATE/03-mobile.yml

@@ -131,6 +131,6 @@ body:
     id: ep-version
     attributes:
       label: Execution Provider Library Version
-      placeholder: ex. CUDA 11.6 or ROCm 5.1.1
+      placeholder: ex. CUDA 12.2 or ROCm 5.1.1
     validations:
       required: false

.github/ISSUE_TEMPLATE/05-performance.yml

@@ -127,7 +127,7 @@ body:
     id: ep-version
     attributes:
       label: Execution Provider Library Version
-      placeholder: ex. CUDA 11.6 or ROCm 5.1.1
+      placeholder: ex. CUDA 12.2 or ROCm 5.1.1
     validations:
       required: false
   - type: textarea

.github/ISSUE_TEMPLATE/06-training.yml

@@ -72,6 +72,6 @@ body:
     id: ep-version
     attributes:
       label: Execution Provider Library Version
-      placeholder: ex. CUDA 11.6 or ROCm 5.1.1
+      placeholder: ex. CUDA 12.2 or ROCm 5.1.1
     validations:
       required: false

.github/ISSUE_TEMPLATE/08-general.yml

@@ -125,6 +125,6 @@ body:
     id: ep-version
     attributes:
       label: Execution Provider Library Version
-      placeholder: ex. CUDA 11.6 or ROCm 5.1.1
+      placeholder: ex. CUDA 12.2 or ROCm 5.1.1
     validations:
       required: false

js/node/script/install.js

@@ -9,7 +9,6 @@
 
 // The purpose of this script is to download the required binaries for the platform and architecture.
 // Currently, most of the binaries are already bundled in the package, except for the following:
-// - Linux/x64/CUDA 11
 // - Linux/x64/CUDA 12
 //
 // The CUDA binaries are not bundled because they are too large to be allowed in the npm registry. Instead, they are

onnxruntime/python/tools/tensorrt/perf/build/build_image.py

@@ -16,7 +16,6 @@
 import sys
 
 TRT_DOCKER_FILES = {
-    "10.9_cuda11.8_cudnn8": "tools/ci_build/github/linux/docker/Dockerfile.ubuntu_cuda11_tensorrt10",
     "10.9_cuda12.8_cudnn9": "tools/ci_build/github/linux/docker/Dockerfile.ubuntu_cuda12_tensorrt10",
     "BIN": "tools/ci_build/github/linux/docker/Dockerfile.ubuntu_tensorrt_bin",
 }

onnxruntime/python/tools/transformers/models/stable_diffusion/README.md

@@ -143,29 +143,11 @@
 
 ### Setup Environment (CUDA) without docker
 
-First, we need install CUDA 11.8 or 12.x, [cuDNN](https://docs.nvidia.com/deeplearning/cudnn/install-guide/index.html), and [TensorRT](https://docs.nvidia.com/deeplearning/tensorrt/install-guide/index.html) in the machine.
+First, we need install CUDA 12.x, [cuDNN](https://docs.nvidia.com/deeplearning/cudnn/install-guide/index.html), and [TensorRT](https://docs.nvidia.com/deeplearning/tensorrt/install-guide/index.html) in the machine.
 
 The verison of CuDNN can be found in https://onnxruntime.ai/docs/execution-providers/CUDA-ExecutionProvider.html#requirements.
 The version of TensorRT can be found in https://onnxruntime.ai/docs/execution-providers/TensorRT-ExecutionProvider.html#requirements.
 
-#### CUDA 11.8:
-
-In the Conda environment, install PyTorch 2.1 up to 2.3.1, and other required packages like the following:
-```
-pip install torch>=2.1,<2.4 --index-url https://download.pytorch.org/whl/cu118
-pip install --upgrade polygraphy onnx-graphsurgeon --extra-index-url https://pypi.ngc.nvidia.com
-pip install -r requirements/cuda11/requirements.txt
-```
-
-For Windows, install nvtx like the following:
-```
-conda install -c conda-forge nvtx
-```
-
-We cannot directly `pip install tensorrt` for CUDA 11. Follow https://github.com/NVIDIA/TensorRT/issues/2773 to install TensorRT for CUDA 11 in Linux.
-
-For Windows, pip install the tensorrt wheel in the downloaded TensorRT zip file instead. Like `pip install tensorrt-8.6.1.6.windows10.x86_64.cuda-11.8\tensorrt-8.6.1.6\python\tensorrt-8.6.1-cp310-none-win_amd64.whl`.
-
 #### CUDA 12.*:
 The official package of onnxruntime-gpu 1.19.x is built for CUDA 12.x. You can install it and other python packages like the following:
 ```

tools/ci_build/github/azure-pipelines/c-api-noopenmp-packaging-pipelines.yml

@@ -76,12 +76,12 @@ variables:
 - name: ReleaseVersionSuffix
   value: ''
 - name: win_trt_version
-  value: 11.8
+  value: 12.2
 
 - name: win_trt_home
-  value: $(Agent.TempDirectory)\${{ variables.win_trt_folder_cuda11 }}
+  value: $(Agent.TempDirectory)\${{ variables.win_trt_folder_cuda12 }}
 - name: win_cuda_home
-  value: $(Agent.TempDirectory)\v11.8
+  value: $(Agent.TempDirectory)\v12.2
 
 stages:
 - template: stages/set_packaging_variables_stage.yml
@@ -110,13 +110,13 @@ stages:
 
 - template: stages/java-cuda-packaging-stage.yml
   parameters:
-    CudaVersion: 11.8
+    CudaVersion: 12.2
     SpecificArtifact: ${{ parameters.SpecificArtifact }}
     BuildId: ${{ parameters.BuildId }}
 
 - template: stages/nuget-combine-cuda-stage.yml
   parameters:
-      CudaVersion: 11.8
+      CudaVersion: 12.2
       RunOnnxRuntimeTests: ${{ parameters.RunOnnxRuntimeTests }}
       UseIncreasedTimeoutForTests: ${{ parameters.UseIncreasedTimeoutForTests }}
       win_trt_home: ${{ variables.win_trt_home }}

tools/ci_build/github/azure-pipelines/cuda-packaging-pipeline.yml

@@ -50,7 +50,6 @@ parameters:
   type: string
   default: '12.2'
   values:
-  - 11.8
   - 12.2
 
 variables:

tools/ci_build/github/azure-pipelines/linux-gpu-ci-pipeline.yml

@@ -33,7 +33,6 @@ parameters:
     type: string
     default: '12.2'
     values:
-      - 11.8
       - 12.2
 
   - name: SpecificArtifact

tools/ci_build/github/azure-pipelines/linux-gpu-tensorrt-ci-pipeline.yml

@@ -33,7 +33,6 @@ parameters:
     type: string
     default: '12.2'
     values:
-      - 11.8
       - 12.2
 
 variables:

tools/ci_build/github/azure-pipelines/linux-gpu-tensorrt-cuda-minimal-ci-pipeline.yml

@@ -33,7 +33,6 @@ parameters:
     type: string
     default: '12.2'
     values:
-      - 11.8
       - 12.2
 
 variables:

tools/ci_build/github/azure-pipelines/linux-gpu-tensorrt-daily-perf-pipeline.yml

@@ -10,7 +10,6 @@ parameters:
   type: string
   default: 10.9_cuda12.8_cudnn9
   values:
-  - 10.9_cuda11.8_cudnn8
   - 10.9_cuda12.8_cudnn9
   - BIN
 

tools/ci_build/github/azure-pipelines/post-merge-jobs.yml

@@ -4,7 +4,6 @@ parameters:
   type: string
   default: '12.2'
   values:
-    - 11.8
     - 12.2
 
 variables:

tools/ci_build/github/azure-pipelines/stages/jobs/py-linux-cuda-package-test-job.yml

@@ -2,9 +2,8 @@ parameters:
   - name: CudaVersion
     displayName: 'CUDA version'
     type: string
-    default: '11.8'
+    default: '12.2'
     values:
-      - 11.8
       - 12.2
   - name: machine_pool
     type: string

tools/ci_build/github/azure-pipelines/stages/nuget-linux-cuda-packaging-stage.yml

@@ -1,7 +1,7 @@
 parameters:
 - name: CudaVersion
   type: string
-  default: '11.8'
+  default: '12.2'
 - name: buildJava
   type: boolean
 - name: buildNodejs

tools/ci_build/github/azure-pipelines/stages/py-gpu-packaging-stage.yml

@@ -36,7 +36,6 @@ parameters:
   displayName: 'CUDA version. Windows Only.'
   default: '12.2'
   values:
-   - 11.8
    - 12.2
 
 - name: PythonVersions

tools/ci_build/github/azure-pipelines/stages/py-linux-gpu-stage.yml

@@ -22,9 +22,8 @@ parameters:
 
 - name: cuda_version
   type: string
-  default: '11.8'
+  default: '12.2'
   values:
-   - 11.8
    - 12.2
 
 stages:
@@ -67,8 +66,8 @@ stages:
 
       - template: ../templates/get-docker-image-steps.yml
         parameters:
-          Dockerfile: tools/ci_build/github/linux/docker/inference/${{ parameters.arch }}/python/cuda/Dockerfile
-          Context: tools/ci_build/github/linux/docker/inference/${{ parameters.arch }}/python/cuda
+          Dockerfile: tools/ci_build/github/linux/docker/inference/${{ parameters.arch }}/default/cuda12/Dockerfile
+          Context: tools/ci_build/github/linux/docker/inference/${{ parameters.arch }}/default/cuda12
           DockerBuildArgs: "--build-arg BASEIMAGE=${{ parameters.docker_base_image }} --build-arg TRT_VERSION=${{ variables.trt_version }} --build-arg BUILD_UID=$( id -u )"
           Repository: onnxruntimecuda${{ replace(parameters.cuda_version, '.', '') }}xtrt86build${{ parameters.arch }}
 

tools/ci_build/github/azure-pipelines/stages/py-win-gpu-stage.yml

@@ -19,9 +19,8 @@ parameters:
   default: ''
 - name: CudaVersion
   type: string
-  default: '11.8'
+  default: '12.2'
   values:
-    - 11.8
     - 12.2
 
 - name: cmake_build_type

tools/ci_build/github/azure-pipelines/templates/jobs/set-winenv.yml

@@ -10,12 +10,12 @@ parameters:
   - name: PrimaryCUDAVersion
     type: string
     default: '12.2'
-  - name: SecondaryCUDAVersion
-    type: string
-    default: '11.8'
-  - name: win_trt_folder_cuda11
-    type: string
-    default: 'TensorRT-10.9.0.34.Windows10.x86_64.cuda-11.8'
+#  - name: SecondaryCUDAVersion
+#    type: string
+#    default: '11.8'
+#  - name: win_trt_folder_cuda11
+#    type: string
+#    default: 'TensorRT-10.9.0.34.Windows10.x86_64.cuda-11.8'
   - name: win_trt_folder_cuda12
     type: string
     default: 'TensorRT-10.9.0.34.Windows10.x86_64.cuda-12.8'
@@ -25,16 +25,16 @@ steps:
       - powershell: |
           azcopy.exe cp --recursive "https://lotusscus.blob.core.windows.net/models/cuda_sdk/v${{ parameters.PrimaryCUDAVersion }}" $(Agent.TempDirectory)
         displayName: 'Download Primary CUDA SDK v${{ parameters.PrimaryCUDAVersion }}'
-      - powershell: |
-          azcopy.exe cp --recursive "https://lotusscus.blob.core.windows.net/models/cuda_sdk/v${{ parameters.SecondaryCUDAVersion }}" $(Agent.TempDirectory)
-        displayName: 'Download Secondary CUDA SDK v${{ parameters.SecondaryCUDAVersion }}'
+#      - powershell: |
+#          azcopy.exe cp --recursive "https://lotusscus.blob.core.windows.net/models/cuda_sdk/v${{ parameters.SecondaryCUDAVersion }}" $(Agent.TempDirectory)
+#        displayName: 'Download Secondary CUDA SDK v${{ parameters.SecondaryCUDAVersion }}'
   - ${{ if eq(parameters.DownloadTRT, 'true') }}:
       - powershell: |
           azcopy.exe cp --recursive "https://lotusscus.blob.core.windows.net/models/local/${{ parameters.win_trt_folder_cuda12 }}" $(Agent.TempDirectory)
         displayName: 'Download ${{ parameters.win_trt_folder_cuda12 }}'
-      - powershell: |
-          azcopy.exe cp --recursive "https://lotusscus.blob.core.windows.net/models/local/${{ parameters.win_trt_folder_cuda11 }}" $(Agent.TempDirectory)
-        displayName: 'Download ${{ parameters.win_trt_folder_cuda11 }}'
+#      - powershell: |
+#          azcopy.exe cp --recursive "https://lotusscus.blob.core.windows.net/models/local/${{ parameters.win_trt_folder_cuda11 }}" $(Agent.TempDirectory)
+#        displayName: 'Download ${{ parameters.win_trt_folder_cuda11 }}'
 
   - task: BatchScript@1
     displayName: 'setup env'