Overview

Semiconductor Data ScienceSimulated · Personal Portfolio

Personal portfolio project showcasing end-to-end fab front-end data science (GCP BigQuery ETL, XGBoost / CatBoost yield prediction, Isolation Forest anomaly detection). Simulated data — not affiliated with or representative of any specific manufacturer.

Overall Yield

92.06%

12,500 wafers processed

Defect Count

45,204

1,589 wafers below 85% yield

Equipment OEE

78.07%

avg utilization across fleet

Active Lots

500

simulated 6-month production window

Yield by Product

Advanced nodes (N3 / N5) show expected lower yield vs mature nodes (N10 / N14).

Yield by Product

Loading from BigQuery…

Defect Type Distribution

Particle defects dominate, consistent with fab contamination patterns.

Defect Type Distribution

Loading from BigQuery…

Equipment Utilization — Top 6

OEE proxy by equipment. MTBF / MTTR from simulated reliability data.

Equipment	Type	Location	Utilization	MTBF (h)	MTTR (h)	Wafers	Avg Yield
EQ-025	Lithography	Bay-02	89.5%	960	3.2	401	92.92%
EQ-018	Etch	Bay-01	88.0%	587	1.9	424	90.85%
EQ-011	Metrology	Bay-06	87.5%	1044	4.0	392	95.21%
EQ-013	Lithography	Bay-02	86.2%	932	6.1	425	92.92%
EQ-006	Etch	Bay-01	86.0%	556	4.5	430	90.79%
EQ-026	CVD	Bay-03	86.0%	974	4.4	411	91.75%

Equipment Utilization — Top 6

Loading from BigQuery…