out-of-core computing

Accelerated Computing CUDA CUDA Programming and Performance
1

Hey folks,

Can anyone give me some suggestions as to how to proceed with out of core computing on gpu. I am working on molecular dynamics and my input data set is large enough that it cannot fit on the GPU. Thus, out-of-core. But I have no idea how to go about it.

Please help

2

Have you tried looking for any literature describing implementations of the algorithm you are using for distributed memory systems?