Su M.-YLin W.-CKuo Y.-TLi C.-MFang E.J.-WHsueh S.S.-Y.CHIEN-MO LI2021-09-022021-09-022021https://www.scopus.com/inward/record.uri?eid=2-s2.0-85106566003&doi=10.1109%2fVLSI-DAT52063.2021.9427338&partnerID=40&md5=941d6f0f6ad68a394d63a357d040f753https://scholars.lib.ntu.edu.tw/handle/123456789/580670Process variation cause a big variation on chip performance, so we need to apply expensive functional test to do the speed binning. In this work, we propose a machine learning-based chip performance prediction framework. We only consider on-chip ring oscillator's frequency as feature, which can be obtained from structural test. We select most important cells for ring oscillators at pre-silicon stage, so we can minimize the ring oscillators on the chip. Experimental results on 12K industry chips show that our prediction accuracy is comparable to automation test equipment's measurement according to company's criterion. ? 2021 IEEE.Automation; Equipment testing; Forecasting; Predictive analytics; Turing machines; VLSI circuits; Automation tests; Chip performance; Functional test; Machine learning techniques; Prediction accuracy; Process Variation; Ring oscillator; Structural tests; Machine learningconference paper10.1109/VLSI-DAT52063.2021.94273382-s2.0-85106566003