GitLab

WCA force now written directly to cuda.force4 array

Reduced memory footprint

Set nonbondThreadsPerBlock to gpu->sim.nonbond_threads_per_block

-This line, and th -se below, will be ignored--

M    amoebaCudaGpu.cpp
M    kCalculateAmoebaCudaMapTorques.cu
M    kCalculateAmoebaCudaRotateFrame.cu

Modified cudaLoadCudaFloatArray() call
Minor optimizations

Added diagnostics for handling reorder of atoms

Bug in Urey-Bradley force calculation fixed

Added support for missing axis types (Cuda only)
Removed rotationMatrix

Renamed files to be consistent w/ other file names

Removal SASA code
Versioning introduced in AmoebaTinkerParameterFile
Decomposition of PME forces included in prm files 

An exception is now thrown if the multipole axis type is not Bisector or Z-then-X
Redid velocity initialization in test code