You may notice very slight differences in the stress and distortion outputs from one run to the next when using more than one core. This is a due to a known limitation in shared-memory parallel processing sometimes referred to as "noise." Operational randomness and numerical round-off inherent to parallelism can cause slightly different results between runs on the same machine using the same number of cores or different numbers of cores. This difference is often negligible. Assuming we compute displacements up to the order of ~1 mm, you may expect the noise to be on the order of ~1e-5 mm. Assuming stress values on the order of ~1 MPa, you may see differences on the order of ~10 Pa. However, in some cases the difference is appreciable. This sort of behavior is most seen when simulations are numerically unstable, that is, not converging. The more numerically unstable the model is, the more likely the convergence pattern or results will differ as the number of cores used in the simulation is changed.

If run with a single thread, the results between successive runs with identical inputs should produce identical outputs.