NVIDIA Certified Professional - Accelerated Data Science

Validates professional-level proficiency in leveraging GPU-accelerated tools and libraries for advanced data science workflows, including data analysis with anomaly detection and time-series analysis, data manipulation and software literacy with ETL and distributed processing, data preparation with cuDF cleansing and synthetic data generation, GPU and cloud computing optimization, machine learning with feature engineering and multi-GPU training, and MLOps practices for benchmarking, memory optimization, and model deployment. The exam covers six domains: Data Manipulation and Software Literacy (19%), MLOps (19%), Data Preparation (17%), GPU and Cloud Computing (16%), Machine Learning (15%), and Data Analysis (14%). Format: 60-70 multiple-choice questions, 120 minutes, proctored online.