GPUEnergyIQ

Graphics Processing Units (GPUs) have become strategic components in high-performance applications such as artificial intelligence and graphics rendering. To support more accurate hardware selection decisions, it is essential to have standardized and up-to-date technical information about GPUs. This paper presents GPUEnergyIQ, a dataset that compiles technical specifications and performance metrics of hundreds of GPU models, collected from public sources such as Kaggle, TechPowerUp, and PassMark. The dataset construction involved stages of data collection, cleaning, standardization, integration, and derivation of variables, including an efficiency metric based on the ratio between performance and energy consumption. GPUEnergyIQ can be applied in comparative analyses of GPUs from different manufacturers, categories, and generations, supporting studies on energy efficiency, technological evolution, and optimal hardware selection in academic, corporate, and high-performance computing contexts.