I use SW2023. 

The model contains one sphere with R=0.001m, m=1kg, with center in the point with coordinates (1,1,1) (in meters).

I've analytically calculated inertia tensor for this model (in R=0 assumption) for coordinate system shown, and I've got

2 -1 -1

-1 2 -1

-1 -1 2 (kg*m2)

While Solidworks gives me another tensor in both positive and negative notation:

1.

Taken at the output coordinate system. (Using