[IntelNav] ci: narrow CUDA arch list to 86;89 for reasonable build time

Previous 75;80;86;89;90 = 5 SASS targets per file × ~200 .cu files
was >60 min even with ccache cold. Consumer coverage maps to just
two arches:
  * sm_86  — Ampere (RTX 30xx, A100-class)
  * sm_89  — Ada Lovelace (RTX 40xx)
Users on Turing / Hopper rebuild locally; we add more back if real
CI demand appears.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

Author: Islam-sehil-59

.github/workflows/intelnav-release.yml

@@ -265,11 +265,17 @@ jobs:
 
       - name: Configure
         run: |
+          # Narrow arch list keeps the CUDA build under the hosted
+          # runner's 6-hour ceiling and within practical ccache reuse.
+          # 86 covers RTX 30-series / A100-like; 89 covers RTX 40.
+          # Users with older (Turing sm_75) or newer (Hopper sm_90)
+          # hardware rebuild from source locally — add them back when
+          # there's real CI demand.
           cmake -B build \
             ${{ env.CMAKE_COMMON }} \
             -DGGML_NATIVE=OFF \
             -DGGML_CUDA=ON \
-            -DCMAKE_CUDA_ARCHITECTURES="75;80;86;89;90" \
+            -DCMAKE_CUDA_ARCHITECTURES="86;89" \
             -DGGML_BACKEND_DL=ON
 
       - name: Build