hooglventures.blogg.se - Opencl benchmark headless

#OPENCL BENCHMARK HEADLESS INSTALL#
#OPENCL BENCHMARK HEADLESS DRIVER#
#OPENCL BENCHMARK HEADLESS VERIFICATION#

the sources are available here: (tag: luxmark_v4.0)

#OPENCL BENCHMARK HEADLESS INSTALL#

Linux 64bit: ( Note: You have to install Qt5 from your distribution in order to run the Linux executable) Windows 64bit: ( Note: you have to install the Microsoft Visual C++ Redistributable for Visual Studio 2017 in order to run the Windows executable)

#OPENCL BENCHMARK HEADLESS VERIFICATION#

ext-info (print scene and image verification too with -single-run) single-run (run the benchmark, print the result to the stdout and exit) mode=BENCHMARK_OCL_GPU|BENCHMARK_OCL_CPUGPU|BENCHMARK_OCL_CPU|BENCHMARK_OCL_CUSTOM|BENCHMARK_HYBRID|BENCHMARK_HYBRID_CUSTOM|BENCHMARK_NATIVE|STRESSTEST_OCL_GPU|STRESSTEST_OCL_CPUGPU|STRESSTEST_OCL_CPU|STRESSTEST_HYBRID|STRESSTEST_NATIVE|DEMO_LUXCOREUI|PAUSE (select the mode to use) scene=FOOD|HALLBENCH|WALLPAPER (select the scene to use) However Intel OpenImageDenoise is a CPU-only task and it is not part of the benchmark result.Ĭommand line options Usage. A comparison between raw and denoised image is shown at the end of the benchmark or during the stress test.

A C++ Bidirectional path tracer with a Metropolis–Hastings sampler.

An hybrid OpenCL for GPUs and C++ for CPUs path tracer with/without Global Illumination cache.

Ray intersection C++ code uses state-of-the-art Intel Embree

A C++ path tracer with/without Global Illumination cache.

A micro-kernel based OpenCL path tracer with/without Global Illumination cache.

This release is based on LuxCoreRender v2.2 and includes benchmarks with the following rendering engines: It is intended as a promotional tool for LuxCoreRender and it is now based on LuxCore API, the LuxCoreRender v2.x C++ or Python API available under Apache Licence v2.0 and freely usable in open source and commercial applications. LuxMark is a OpenCL cross-platform benchmark tool and has become, over past years, one of the most used (if not the most used) OpenCL benchmark.

1.5.3 Bidirectional path tracing with Metropolis sampler (C++ only).

1.5.2 Path tracing with GI cache (OpenCL, C++ and Hybrid).

1.5.1 Brute force path tracing (OpenCL, C++ and Hybrid).

With 20.40 there's no need to disable the option, and I get 30 MRays/sec, and a score of 5800. ROC, and now 20.45 as well, will render the benchmark LuxMark 3.1 incorrectly if compiler option cl-fast-relaxed-math is enabled.Įven if you disable the option, ROC will run quite a bit slower, I get 15 MRays/sec, and a score of 2800.

#OPENCL BENCHMARK HEADLESS DRIVER#

Unfortunately this means that this new driver has inherited an issue that occurs with ROC. I understand a new feature of 20.45 is the backend, it's no longer PAL, it's now ROC. To add insult to injury, because the module fails to build, it will blacklist amdgpu, causing Ubunto to use the wrong driver, and screen resolution on reboot! However it does install OpenCL, so after it has failed I remove amdgpu-dkms, and the blacklist. This is working well with amdgpu-pro 20.40, but 20.45 is just ignoring the -no-dkms part, and tries to create, and install the module anyway. That's why I'm installing it like so: amdgpu-install -no-dkms -headless -opencl=rocr Because I'm installing OpenCL on kernel 5.8, I don't want to build kernel module amdgpu, as this will fail. I'm installing the OpenCL driver from amdgpu-pro on Ubuntu 20.10.