Add support for Pascal and Volta

Author: robballantyne

.github/workflows/build-cuda.yml

@@ -63,23 +63,23 @@ jobs:
           - cuda_version: '12.4.1'
             cuda_version_short: '12.4'
             cuda_tag: '12.4.1-cudnn-devel-ubuntu22.04'
-            architectures: '75-virtual;80-virtual;86-virtual;89-virtual;90-virtual'
+            architectures: '61-virtual;70-virtual;75-virtual;80-virtual;86-virtual;89-virtual;90-virtual'
           - cuda_version: '12.6.3'
             cuda_version_short: '12.6'
             cuda_tag: '12.6.3-cudnn-devel-ubuntu22.04'
-            architectures: '75-virtual;80-virtual;86-virtual;89-virtual;90-virtual'
+            architectures: '61-virtual;70-virtual;75-virtual;80-virtual;86-virtual;89-virtual;90-virtual'
           - cuda_version: '12.8.1'
             cuda_version_short: '12.8'
             cuda_tag: '12.8.1-cudnn-devel-ubuntu22.04'
-            architectures: '75-virtual;80-virtual;86-virtual;89-virtual;90-virtual;100-virtual;120-virtual'
+            architectures: '61-virtual;70-virtual;75-virtual;80-virtual;86-virtual;89-virtual;90-virtual;100-virtual;120-virtual'
           - cuda_version: '12.9.1'
             cuda_version_short: '12.9'
             cuda_tag: '12.9.1-cudnn-devel-ubuntu22.04'
-            architectures: '75-virtual;80-virtual;86-virtual;89-virtual;90-virtual;100-virtual;120-virtual'
+            architectures: '61-virtual;70-virtual;75-virtual;80-virtual;86-virtual;89-virtual;90-virtual;100-virtual;120-virtual'
           - cuda_version: '13.0.1'
             cuda_version_short: '13.0'
             cuda_tag: '13.0.1-cudnn-devel-ubuntu22.04'
-            architectures: '75-virtual;80-virtual;86-virtual;89-virtual;90-virtual;100-virtual;120-virtual'
+            architectures: '61-virtual;70-virtual;75-virtual;80-virtual;86-virtual;89-virtual;90-virtual;100-virtual;120-virtual'
 
     steps:
       - name: Checkout
@@ -240,14 +240,16 @@ jobs:
             **Commit:** ${{ needs.check-release.outputs.release_hash }}
             
             ## CUDA Versions
-            - CUDA 12.4 - Architectures: 7.5, 8.0, 8.6, 8.9, 9.0
-            - CUDA 12.6 - Architectures: 7.5, 8.0, 8.6, 8.9, 9.0
-            - CUDA 12.8 - Architectures: 7.5, 8.0, 8.6, 8.9, 9.0, 10.0, 12.0
-            - CUDA 12.9 - Architectures: 7.5, 8.0, 8.6, 8.9, 9.0, 10.0, 12.0
-            - CUDA 13.0 - Architectures: 7.5, 8.0, 8.6, 8.9, 9.0, 10.0, 12.0
+            - CUDA 12.4 - Architectures: 6.1, 7.0, 7.5, 8.0, 8.6, 8.9, 9.0
+            - CUDA 12.6 - Architectures: 6.1, 7.0, 7.5, 8.0, 8.6, 8.9, 9.0
+            - CUDA 12.8 - Architectures: 6.1, 7.0, 7.5, 8.0, 8.6, 8.9, 9.0, 10.0, 12.0
+            - CUDA 12.9 - Architectures: 6.1, 7.0, 7.5, 8.0, 8.6, 8.9, 9.0, 10.0, 12.0
+            - CUDA 13.0 - Architectures: 6.1, 7.0, 7.5, 8.0, 8.6, 8.9, 9.0, 10.0, 12.0
             
             ## Architecture Reference
-            - 7.5: Tesla T4, RTX 2000 series, Quadro RTX
+            - 6.1: Titan XP, Tesla P40, GTX 10xx
+            - 7.0: Tesla V100
+            - 7.5: Tesla T4, RTX 20xx series, Quadro RTX
             - 8.0: A100
             - 8.6: RTX 3000 series
             - 8.9: RTX 4000 series, L4, L40

README.md

@@ -24,12 +24,14 @@ The official llama.cpp repository does not provide pre-built CUDA binaries. This
 
 | Compute Capability | GPU Examples | CUDA 12.4/12.6 | CUDA 12.8+ |
 |-------------------|--------------|----------------|------------|
+| 6.1 | Titan XP, Tesla P40, GTX 10xx | ✅ | ✅ |
+| 7.0 | Tesla V100 | ✅ | ✅ |
 | 7.5 | Tesla T4, RTX 2000 series, Quadro RTX | ✅ | ✅ |
 | 8.0 | A100 | ✅ | ✅ |
 | 8.6 | RTX 3000 series | ✅ | ✅ |
 | 8.9 | RTX 4000 series, L4, L40 | ✅ | ✅ |
 | 9.0 | H100, H200 | ✅ | ✅ |
-| 10.0 | B100, B200, GB200 | ❌ | ✅ |
+| 10.0 | B200 | ❌ | ✅ |
 | 12.0 | RTX Pro series, RTX 5000 series | ❌ | ✅ |
 
 **Note:** Blackwell architectures (10.0 & 12.0) require CUDA 12.8 or later.