sum everything to a single element with threads

[compact portal]

sum everything to a single element with threads

[glass grotto]

This is trivial task for compute shader:

StructuredBuffer<int> valuesArray;
RWStructuredBuffer<int> resultInFirstElement;

[numthreads(256, 1, 1)]
void ParallelSum(uint tid: SV_DispatchThreadID)
{
    InterlockedAdd(resultInFirstElement[0], valuesArray[tid]);
}