Historique | Voir | Annoter | Télécharger (31,35 ko)
Add LogItops as output.
Clean some extra printing.
Add option on Seeds in order to improve // implementations under Threads and MPI
Set Seed as static number to evaluate reproducibility.
Change Compile code to match Stretch evolution
Define the last value of Parallel Rate exploration as the first one if not defined.
Add -k option to perform classical test based on if.
Correct CUDA call with direct ccbin compiler call (clang compiles but gcc over 5 not...)
Replace *PU by xPU to avoid prolems on files.
Add C/OpenCL implementation.
Modify the "device.type" in order not to crash with POCL implementation.
Add Pi estimation.
Transform from Python 2 to Python 3
Identation for XPU parsing.
Correct CUDA implementation.
Add MPI and pThreads version to use multiple OpenCL devices.
Add MPI version which support multi GPU.
Change structure of call for OpenCL Metropolis Routine
Internal modifications on statistics.
Changes on micmac global variables.
Minor change on doc.