Demonstates how to load and execute a CUDA kernel in OpenInventor.
This demo loads a simple CUDA kernel using InventorCUDA API. The kernel is executed and the back is read back for comparaison.
The comparaison code does a floating error computation and display it in the console.