Repository Name Description Repository Link Communities
CUDA SDK Wrapper Library

The CUDA SDK wrapper library provides means for an efficient resource sharing and resource protection on multi-user GPU clusters, such as NCSA's 32-node 128-GPU system. It implements the following functionality:
  • Virtualization of the physical GPU devices. The virtual devices visible to the user map to a consistent set of physical devices, which accomplishes "user fencing" on shared systems and prevents users from accidentally trampling one another.
  • Rotation of the virtual to physical mapping for each new process that requests a GPU resource. This provides a method for large parallel tasks to use common startup parameters and still use multiple device targets. I.e., when each new process calls for gpu0, the underlying physical device gets shifted allowing for the next process calling for gpu0 to get the next allocated physical device.
  • Ensuring NUMA affinity for GPUs on systems that have multiple memory controllers. NUMA affinity can be mapped between CPU cores and GPU devices. This has been shown to have as much as 6-8% improvement in host to device memory bandwidth.
  • Memory-scrubbing to wipe the user's GPU memory after use for security from subsequent users.
When installed, the CUDA SDK wrapper library is forced preload and intercepts the device allocation calls to CUDA libraries in order to provide the above-mentioned functionality.
 http://sourceforge.net/projects/cudawrapper/ Building High Performance Computers and Clusters with GPU's, GPU Computing General Topics
CUDA Memory Test

 The CUDA GPU Memory Test adopts memory test methodology used in Memtest86 utility for GPU device memory. The main idea of the memory test utility is to write a test pattern to memory, read it back and verify if it is the same as written, write the pattern's complement to the memory, read and verify it again. All 10 tests from Memtest86, plus one additional test are implemented. These tests are designed to catch both permanent hardware errors due to manufacturing defects and prolonged use of memory chips and "soft errors" due to cosmic radiation. The GPU memory test can be continuously used to monitor the system for appearance of new hardware faults and to collect statistics about the rate of soft errors.

 http://sourceforge.net/projects/cudagpumemtest/ Building High Performance Computers and Clusters with GPU's, GPU Computing General Topics
Graphics Gems Repository

This is the official on-line repository for the code from the Graphics Gems series of books (from Academic Press). This series focusses on short to medium length pieces of code which perform a wide variety of computer graphics related tasks. All code here can be used without restrictions. The code distributions here contain all known bug fixes and enhancements. We also provide errata listings for the text of each book.

The gems can be viewed by category, by book, or by author. Gems code can be accessed in a variety of ways: by viewing the code directly from these pages or by downloading the whole website or a book's entire code base.  

 http://www.graphicsgems.org/ GPU Computing General Topics
SIML: A Fast SIMD Implementation of LINGO Chemical Similarities

SIML ("Single-Instruction, Multiple-LINGO") is a library containing implementations of a fast SIMD algorithm for calculating the LINGO chemical similarity metric. This method, currently implemented for CUDA and OpenCL-supporting GPUs is up to two orders of magnitude faster when run on a GPU than existing commercial implementations of LINGO.

This repository is associated with the GPU Computing Gems, vol. 1 chapter "Large-Scale Chemical Informatics on GPUs" and the publication "SIML: A Fast SIMD Algorithm for Calculating LINGO Chemical Similarities on GPUs and CPUs" (http://pubs.acs.org/doi/abs/10.1021/ci100011z).

 https://simtk.org/home/siml GPU Computing Gems Source Code
PAPER Accelerates Parallel Evaluations of ROCS

PAPER is a program to calculate optimal overlays of and similarities between 3-D representations of molecules, based on the Gaussian Shape Overlay algorithm (as used, for example, in OpenEye ROCS). It accelerates large screening experiments by evaluating multiple overlays in parallel on NVIDIA GPUs.

This repository is associated with the GPU Computing Gems, vol. 1 chapter "Large-Scale Chemical Informatics on GPUs" and the publication "PAPER - Accelerating Parallel Evaluations of ROCS" (http://dx.doi.org/10.1002/jcc.21307).

 https://simtk.org/home/paper GPU Computing Gems Source Code
Lattice-Boltzmann Lighting Models - Source Code

The link points to a .tar.gz code collection that contains:

1. OpenCL code for lattice-Boltzmann lighting of participating media.

2. A sample cloud density file.

3. A C code converter to build a voxel file from the density file and its lighting solution.

4. A C code, interactive voxel viewer.

See the README file for instructions.

 http://www.cs.clemson.edu/~geist/lblightingcode.html GPU Computing Gems Source Code