Conversation
|
✅ Результаты тестирования PR #1022 Логи тестирования (нажмите чтобы развернуть)=== СТАТУС: Успешно выполнены программы: main_linear_bvh === === main_linear_bvh stdout (exit code: -11 (segfault после выполнения)) === Found 1 GPUs in 8.98923 sec (CUDA: 0.11636 sec, OpenCL: 1.35557 sec, Vulkan: 7.51724 sec) Available devices: Device #0: API: CUDA+OpenCL+Vulkan. GPU. Tesla T4 (CUDA 12020). Free memory: 14822/14930 Mb. Using device #0: API: CUDA+OpenCL+Vulkan. GPU. Tesla T4 (CUDA 12020). Free memory: 14822/14930 Mb. Using OpenCL API... Using 8 ray samples for ambient occlusion ____________________________________________________________________________________________ Loading scene data/gnome/gnome.ply... Loading camera data/gnome/camera.txt... Scene gnome loaded: 1297 vertices, 764 faces in 0.00247211 sec Camera framebuffer size: 2177x1901 Kernels compilation done in 3.62687 seconds GPU brute force ray tracing frame render times (in seconds) - 10 values (min=0.3252 10%=0.325348 median=0.326654 90%=4.09166 max=4.09166) CPU build LBVH in 0.000262601 sec CPU LBVH build performance: 2.90936 MTris/s Kernels compilation done in 0.756482 seconds GPU with CPU LBVH ray tracing frame render times (in seconds) - 10 values (min=0.0647812 10%=0.064833 median=0.0651361 90%=0.813428 max=0.813428) GPU with CPU LBVH ray tracing performance: 508.287 MRays/s Scene processed in 10.0326 sec = 0% scene IO + 70% brute force RT + 0% CPU LBVH + 14% GPU with CPU LBVH + 0% GPU LBVH + 0% GPU with GPU LBVH + 15% images IO + 0% PCI-E write + 0% PCI-E read + 0% cleaning VRAM ____________________________________________________________________________________________ Loading scene data/powerplant/powerplant.obj... Loading camera data/powerplant/camera.txt... Scene powerplant loaded: 5984083 vertices, 12759246 faces in 1.85114 sec Camera framebuffer size: 3228x1901 CPU build LBVH in 5.88712 sec CPU LBVH build performance: 2.16731 MTris/s GPU with CPU LBVH ray tracing frame render times (in seconds) - 10 values (min=1.03984 10%=1.03994 median=1.04046 90%=1.13202 max=1.13202) GPU with CPU LBVH ray tracing performance: 47.1825 MRays/s Scene processed in 19.6577 sec = 9% scene IO + 30% CPU LBVH + 53% GPU with CPU LBVH + 0% GPU LBVH + 0% GPU with GPU LBVH + 6% images IO + 0% PCI-E write + 0% PCI-E read + 1% cleaning VRAM ____________________________________________________________________________________________ Loading scene data/san-miguel/san-miguel.obj... Loading camera data/san-miguel/camera.txt... Scene san-miguel loaded: 3738829 vertices, 5617451 faces in 1.08474 sec Camera framebuffer size: 3840x2108 CPU build LBVH in 2.62405 sec CPU LBVH build performance: 2.14076 MTris/s GPU with CPU LBVH ray tracing frame render times (in seconds) - 10 values (min=2.17531 10%=2.17704 median=2.186 90%=2.24494 max=2.24494) GPU with CPU LBVH ray tracing performance: 29.6239 MRays/s Scene processed in 27.2383 sec = 4% scene IO + 10% CPU LBVH + 80% GPU with CPU LBVH + 0% GPU LBVH + 0% GPU with GPU LBVH + 6% images IO + 0% PCI-E write + 0% PCI-E read + 0% cleaning VRAM ____________________________________________________________________________________________ Total GPU RT with LBVH avg perf: 195.031 MRays/sec (all 3 values: [508.287 47.1825 29.6239]) Total building GPU LBVH avg perf: 0 MTris/sec (all 0 values: []) Final score: 0 coolness Results are incomplete! |
Логи тестирования (нажмите чтобы развернуть)Ошибка компиляции |
|
✅ Результаты тестирования PR #1022 Логи тестирования (нажмите чтобы развернуть)=== СТАТУС: Успешно выполнены программы: main_linear_bvh === === main_linear_bvh stdout (exit code: -3 (таймаут, но вывод получен)) === Found 1 GPUs in 8.64152 sec (CUDA: 0.114962 sec, OpenCL: 0.706186 sec, Vulkan: 7.8203 sec) Available devices: Device #0: API: CUDA+OpenCL+Vulkan. GPU. Tesla T4 (CUDA 12020). Free memory: 14822/14930 Mb. Using device #0: API: CUDA+OpenCL+Vulkan. GPU. Tesla T4 (CUDA 12020). Free memory: 14822/14930 Mb. Using OpenCL API... Using 8 ray samples for ambient occlusion ____________________________________________________________________________________________ Loading scene data/gnome/gnome.ply... Loading camera data/gnome/camera.txt... Scene gnome loaded: 1297 vertices, 764 faces in 0.00256114 sec Camera framebuffer size: 2177x1901 Kernels compilation done in 3.47323 seconds GPU brute force ray tracing frame render times (in seconds) - 10 values (min=0.312727 10%=0.313844 median=0.315396 90%=3.88145 max=3.88145) CPU build LBVH in 0.000273342 sec CPU LBVH build performance: 2.79503 MTris/s Kernels compilation done in 0.141757 seconds GPU with CPU LBVH ray tracing frame render times (in seconds) - 10 values (min=0.0619113 10%=0.0619166 median=0.0623122 90%=0.198129 max=0.198129) GPU with CPU LBVH ray tracing performance: 531.321 MRays/s Kernels compilation done in 0.51888 seconds Kernels compilation done in 0.115396 seconds Kernels compilation done in 0.176926 seconds Kernels compilation done in 0.115325 seconds Kernels compilation done in 0.124996 seconds GPU LBVH build times (in seconds) - 10 values (min=0.00279994 10%=0.00280558 median=0.00281745 90%=1.05468 max=1.05468) GPU LBVH build performance: 0.271167 MTris/s GPU with GPU LBVH ray tracing frame render times (in seconds) - 10 values (min=0.0564658 10%=0.0566896 median=0.0574181 90%=0.124719 max=0.124719) GPU with GPU LBVH ray tracing performance: 576.609 MRays/s Scene processed in 11.6564 sec = 0% scene IO + 58% brute force RT + 0% CPU LBVH + 7% GPU with CPU LBVH + 9% GPU LBVH + 6% GPU with GPU LBVH + 20% images IO + 0% PCI-E write + 0% PCI-E read + 0% cleaning VRAM ____________________________________________________________________________________________ Loading scene data/powerplant/powerplant.obj... Loading camera data/powerplant/camera.txt... Scene powerplant loaded: 5984083 vertices, 12759246 faces in 1.86322 sec Camera framebuffer size: 3228x1901 CPU build LBVH in 5.96121 sec CPU LBVH build performance: 2.14038 MTris/s GPU with CPU LBVH ray tracing frame render times (in seconds) - 10 values (min=1.00481 10%=1.00568 median=1.00639 90%=1.08636 max=1.08636) GPU with CPU LBVH ray tracing performance: 48.7795 MRays/s |
|
пожалуйста протестируйте на всех данных |
3 participants
Локальный вывод
Вывод Github CI