If this is your first visit, be sure to
check out the FAQ by clicking the
link above. You may have to register
before you can post: click the register link above to proceed. To start viewing messages,
select the forum that you want to visit from the selection below.
Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite
Available on the Command Line (All Platforms) and Windows GUI. (Mac / Linux GUI's will come in a later release) Currently only available in OpenCL form so requires a AMD or Intel GPU supporting OpenCL 1.1 or later. Nvidia GPU's are not currently supported.
When downscaling, up to 5% performance improvement can be achieved. No benefit when not downscaling.
Small loss in quality over the Lanczos scaler.
What part of Device Agnostic do developers not grasp about OpenCL? This should be available on any OpenCL 1.1 or newer AMD CPU or Intel CPU, not to mention a system that has other OpenCL devices.
What part of Device Agnostic do developers not grasp about OpenCL? This should be available on any OpenCL 1.1 or newer AMD CPU or Intel CPU, not to mention a system that has other OpenCL devices.
The part where every GPU vendor implements OpenCL in a different way and your code doesn't always work as expected in their platforms? 1.1 should work on nVidia, but thay've got to have a reason for saying it's unsupported.
What part of Device Agnostic do developers not grasp about OpenCL? This should be available on any OpenCL 1.1 or newer AMD CPU or Intel CPU, not to mention a system that has other OpenCL devices.
Here is one example: Last time I used it, Nvidia's implementation does not provide the C++ bindings for OpenCL. Probably because it makes OpenCL more comfortable to use than Cuda. So you can't build an OpenCL application based on C++.
Since C++ bindings are one million times better in every way than the standard C interface, programmers like to use them. They really make OpenCL a breeze. Nvidia is the only vendor that got this wrong.
You can get things to work with Nvidia, but you have to drop cl.hpp from some other implementation into your /usr/include and this is not something anybody wants to recommend or support.
Disclaimer -- I don't know if this is the problem Hanbrake is having, but it was the problem I was having. Getting people to compile my code on Nvidia was difficult.
Comment