Folding Forum

Hey guys .

in One of vijay's blogposts he talks about progress of the AMD opencl based folding client and links to a site stating that there is a changelog there
i cant find it, most of the stuff on the site has a datestamp from 2010 or earlier , so i dont see where i would find any recent progress.

does anyone else know where this changelog with recent progress can be found?

Perhaps he was referring to https://simtk.org/project/xml/news.xml?group_id=352 or the new v7 client changelog at http://folding.typepad.com/news/2011/03 ... lease.html

HendricksSA wrote:Perhaps he was referring to https://simtk.org/project/xml/news.xml?group_id=352 or the new v7 client changelog at http://folding.typepad.com/news/2011/03 ... lease.html

Neither. There should be some specific OpenMM changelog somewhere.....

Ivoshiee wrote:

HendricksSA wrote:Perhaps he was referring to https://simtk.org/project/xml/news.xml?group_id=352 or the new v7 client changelog at http://folding.typepad.com/news/2011/03 ... lease.html

Neither. There should be some specific OpenMM changelog somewhere.....

Perhaps this one?
https://simtk.org/project/xml/news.xml? ... =161#news1

Note that like other sites that use https: you must register at simtk.org to see 99% of their content.

The OpenMM code now supports accelerated calculations on NVIDIA GPU cards (using either CUDA or OpenCL), as well as ATI GPU cards (using OpenCL).

The OpenMM code has been ported to FAH and FahCore_15 uses that code to provide CUDA support on NVIDIA.
FahCore_16 is being developed which will be using the OpenMM code to provide OpenCL support for ATI.
(No word yet on whether FAH will support OpenCL for NVIDIA.)

bruce wrote:

Ivoshiee wrote:

HendricksSA wrote:Perhaps he was referring to https://simtk.org/project/xml/news.xml?group_id=352 or the new v7 client changelog at http://folding.typepad.com/news/2011/03 ... lease.html

Neither. There should be some specific OpenMM changelog somewhere.....

Perhaps this one?
https://simtk.org/project/xml/news.xml? ... =161#news1

Note that like other sites that use https: you must register at simtk.org to see 99% of their content.

The OpenMM code now supports accelerated calculations on NVIDIA GPU cards (using either CUDA or OpenCL), as well as ATI GPU cards (using OpenCL).

The OpenMM code has been ported to FAH and FahCore_15 uses that code to provide CUDA support on NVIDIA.
FahCore_16 is being developed which will be using the OpenMM code to provide OpenCL support for ATI.
(No word yet on whether FAH will support OpenCL for NVIDIA.)

if that is the changelog , then there hasnt been any progress on the GPU3 client since September 15th 2010 , ( as far as i can see )
and since Vijay's blogpost was from march 15th 2011 , i was really hoping to see changelog entires atleast that fresh.
it must be somewhere else.

Here's some more info. The changes are to the *core* (the client isn't relevant for these issues). Core16 uses the latest openMM code. Core change info in terms of the science and GPU support (eg AMD support) is kept on the openMM site.