You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
 
 
 
fxzjshm a64594d4b8
Enable BOOST_COMPUTE_USE_OFFLINE_CACHE
4 months ago
example Add argument to test different devices in example/testdedisp.c 10 months ago
html Ran doxygen. 10 years ago
src Fix caching not working 4 months ago
src.orig Move original src 10 months ago
.gitignore Migrate `transpose_kernel` to OpenCL 10 months ago
Doxyfile Initial code commit. Contains working dedisp library source and documentation. 10 years ago
INSTALL Made changes to library and docs regarding the sign of the df parameter. Fixed minor doc and Makefile errors regarding DEDISP_DEBUG mode. 10 years ago
Makefile Enable BOOST_COMPUTE_USE_OFFLINE_CACHE 4 months ago
Makefile.inc Replace OpenCL C++ header include path with OpenCL C header in Makefile.inc 10 months ago
README.md Update README.md 10 months ago
libdedisp.version Added dedisp_get_max_delay 4 years ago

README.md

Migrated to OpenCL using Boost.Compute
Results of example/testdedisp seems passable compared to original CUDA version, though not tested seriously.

Installation: just make && make install it. If it can't be built, check CL_INCLUDE_DIR and BOOST_COMPUTE_INLCUDE_DIR in Makefile.inc

Below is the original Readme.


dedisp

This repositry is derived from Ben Barsdell's original GPU De-dedispersion library (code.google.com/p/dedisp)

Installation Instructions:

  1. git clone https://github.com/ajameson/dedisp.git
  2. Update Makefile.inc with your CUDA path, Install Dir and GPU architecture. e.g.
    • CUDA_PATH ?= /usr/local/cuda-8.0.61
    • INSTALL_DIR = $(HOME)/opt/dedisp
    • GPU_ARCH = sm_60
  3. make && make install

This will build a shared object library named libdedisp.so which is a prerequisite for Heimdall. The dedisp header files will be installed into INSTALL_DIR/include and the library into INSTALL_DIR/lib.