Accurate determination of a chemical mechanical planarization (CMP) process reaching the end point is an important problem in the CMP process. In present work, the variation of the motor power signal of the polishing platen in the CMP process has been investigated, and the moving average method with 121-points moving window was used to smooth the original signal curve. Experiment results showed that the processed signal could facilitate the extraction of end-point feature for the in-situ end-point detection (EPD) system and it was relatively steady before and after the layer transition stage, which made it more reliable to detect the end point in situ. Comparing with other EPD methods, the system was less complicated, and it was easier to code for algorithm development. Finally, further analysis was performed and series of experiments provided a planarized via-revealed surface with low via dishing.

The traditional distributed operating system (OS) and homogeneous multiprocessors OS are not suitable for heterogeneous multi-core processor (HMP) in real-time and efficiency performance. After researching the structure and the development trend of the HMP, this paper presents a multi-master model real-time operating system architecture for HMP, which is designed with symmetric structure and functional modularization, and every kernel can manage resources, tasks and other shared devices as a master to improve the real-time and efficiency performance, to solve the bottle-neck problem that the single master can not satisfy the requirement of performance with increasing of the slave number. This architecture can adapt to different structure processor core with configurable function module. On the other hand, it can also reduce the development difficulty of the heterogeneous multi-core processor OS.