ci: bump latest torch test version to 2.11.0

Update PR and nightly test matrices to include torch 2.11.0 (cu130),
replacing 2.10.0 as the latest tested version.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: Titus-von-Koeller

.github/workflows/tests-nightly.yml

@@ -20,7 +20,7 @@ jobs:
         platform: [linux-x64, linux-aarch64, macos, windows]
         # default runners don't have AVX-512 support, but icelake does
         cpu_type: ["", icelake]
-        torch_version: ["2.3.1", "2.9.1", "2.10.0"]
+        torch_version: ["2.3.1", "2.10.0", "2.11.0"]
 
         exclude:
           # aarch64 minimum torch version is 2.5.1
@@ -71,7 +71,7 @@ jobs:
             torch_version: "2.9.1"
             pypi_index: "https://download.pytorch.org/whl/cu128"
           - cuda_version: "13.0.2"
-            torch_version: "2.10.0"
+            torch_version: "2.11.0"
             pypi_index: "https://download.pytorch.org/whl/cu130"
 
           # Windows CUDA Tests - T4 GPU (CUDA 11.8 only, multiple torch versions)

.github/workflows/tests-pr.yml

@@ -31,7 +31,7 @@ jobs:
         platform: [linux-x64, linux-aarch64, macos]
         # default runners don't have AVX-512 support, but icelake does
         cpu_type: ["", icelake]
-        torch_version: ["2.3.1", "2.10.0"]
+        torch_version: ["2.3.1", "2.11.0"]
 
         exclude:
           # aarch64 minimum torch version is 2.5.1
@@ -76,7 +76,7 @@ jobs:
             torch_version: "2.9.1"
             pypi_index: "https://download.pytorch.org/whl/cu128"
           - cuda_version: "13.0.2"
-            torch_version: "2.10.0"
+            torch_version: "2.11.0"
             pypi_index: "https://download.pytorch.org/whl/cu130"
 
           # Windows CUDA test - single configuration