Releases: fahbench/fahbench
FAHBench 2.3
How fast is your GPU? FAHBench runs a sample molecular dynamics calculation to benchmark GPUs as it pertains to biophysics research. FAHBench performs the same computations as those found on Folding@Home.
This version of FAHBench contains some UI fixes and upgrades OpenMM to be consistent with core21 v0.0.18.
FAHBench 2.2
How fast is your GPU? FAHBench runs a sample molecular dynamics calculation to benchmark GPUs as it pertains to biophysics research. FAHBench performs the same computations as those found on Folding@Home.
This version of FAHBench has a feature-complete GUI as well as a command line interface. It was built with AMD app sdk v2.9-1, openmm 6.2 core 21 v0.0.17.
Installing
FAHBench comes in either a standalone archive (zip
or tar.gz
file) or a graphical installer.
Standalone
Extract the archive and run bin/FAHBench-gui
or bin/FAHBench-cmd
(for graphical interface or command-line interface, resp.)
Windows Installer
Follow the prompts. It will install a FAHBench shortcut to the start menu
On Windows, if you see something like api-ms-win-crt-runtime-l1-1-0.dll missing
, please make sure you've installed all recommended windows updates or install https://www.microsoft.com/en-us/download/details.aspx?id=48234 directly.
Linux
Linux users will need qt5 libraries installed to run the GUI. On debian/ubuntu run sudo apt-get install libqt5widgets5
FAHBench 2.1 with OpenMM 6.2
This release makes CUDA a runtime dependency and fixes some packaging issues. For a full list of changes, check out the 2.1 milestone on github
Known issues:
- The GUI is not feature-complete. There are still some features missing.
Optional Dependencies:
- Linux: Qt5 (GUI), Cuda 6.0 (CUDA)
- Windows: Cuda 6.5 (CUDA)
FAHBench 2.0 with OpenMM 6.2
With this release, FAHBench
- Uses the latest release of OpenMM (v6.2) corresponding to Folding@Home Core 21
- Supports both Windows and Linux
- Has a modular design which will allow future improvements
- Can be compiled and run without a graphical interface (without Qt)
There are still some features missing from the GUI that will be added going forwards. The binary packages may not provide all the necessary dependencies. They may not (but probably should) work on systems other than:
- Windows 8 with cuda 6.5
- Ubuntu 14.04 with cuda 6.0
Known issues:
- The GUI fails to run in windows due to a packaging error. Sorry!
v1.2.1
Update README.md