diff --git a/.github/workflows/nv-accelerate-v100.yml b/.github/workflows/nv-accelerate-v100.yml
new file mode 100644
index 000000000000..20743dc6d858
--- /dev/null
+++ b/.github/workflows/nv-accelerate-v100.yml
@@ -0,0 +1,58 @@
+name: nv-accelerate-v100
+
+on:
+  push:
+    branches:
+      - 'master'
+      - 'staging**'
+    paths-ignore:
+      - 'docs/**'
+  pull_request:
+    paths-ignore:
+      - 'docs/**'
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  unit-tests:
+    runs-on: [self-hosted, nvidia, cu111, v100]
+
+    steps:
+      - uses: actions/checkout@v2
+
+      - name: environment
+        run: |
+          nvidia-smi
+          which python
+          python --version
+          which nvcc
+          nvcc --version
+          pip install --upgrade pip
+          pip uninstall --yes torch torchvision
+          pip install torch==1.8.2+cu111 torchvision==0.9.2+cu111 -f https://download.pytorch.org/whl/lts/1.8/torch_lts.html
+          python -c "import torch; print('torch:', torch.__version__, torch)"
+          python -c "import torch; print('CUDA available:', torch.cuda.is_available())"
+
+      - name: Python environment
+        run: |
+          pip list
+
+      - name: Install deepspeed
+        run: |
+          pip uninstall --yes deepspeed
+          pip install .[dev,autotuning]
+          ds_report
+
+      - name: HF Accelerate tests
+        run: |
+          if [[ -d ./torch-extensions ]]; then rm -rf ./torch-extensions; fi
+          git clone https://github.com/huggingface/accelerate
+          cd accelerate
+          # installing dependencies
+          pip install .[testing]
+          # force protobuf version due to issues
+          pip install "protobuf<4.21.0"
+          pip list
+          TORCH_EXTENSIONS_DIR=./torch-extensions pytest --color=yes --durations=0 --verbose tests/deepspeed