I call this SIMS vs SIMS, because of an earlier page called SIMS vs SRP.

, Charles Evans and Associates, graciously supplied this image. Contact him if you questions about SIMS for TCAD calibration.

This figure highlights the different results that can be obtained on different types of SIMS machines and under different profiling conditions. The quadrupole machine used lower beam energy, and thus did a better job of profiling the channeling tail (less knock-on). However, the magnetic sector instrument did a better job of profiling the surface. You almost need to run both types of machines to get the complete profile (that's my interpretation, not necessarily Dr. Chia's). Any TCAD vendor applications engineer has faced the question "Why doesn't the simulator agree with the measurement", with the implicit assumption that the measurement is always correct. This is a great example of how SIMS is not the perfect reference it is often assumed to be.