NVIDIA Certified Associate - Generative AI Multimodal

Validates foundational skills needed to design, implement, and manage AI systems that synthesize and interpret data across text, image, and audio modalities using NVIDIA platforms. Covers experimentation with multimodal models, core machine learning and AI knowledge including transformer and diffusion architectures, multimodal data processing across text/image/audio/video, software development for multimodal applications, data analysis and visualization, performance optimization for inference and training, and trustworthy AI practices. The exam covers seven domains: Experimentation (25%), Core Machine Learning and AI Knowledge (20%), Multimodal Data (15%), Software Development (15%), Data Analysis and Visualization (10%), Performance Optimization (10%), and Trustworthy AI (5%). Format: 50-60 multiple-choice questions, 60 minutes, proctored online.