Architectural Characteristics of Apple Silicon for Machine Learning Applications; The Rise of Generative Artificial Intelligence as a Technological System

As artificial intelligence systems are increasingly scaled and integrated into society, their sustainability—both in terms of technical infrastructure and societal impact—is being brought into focus. This portfolio reflects a desire to understand and improve the foundations of AI, which is approached from both sociotechnical and system-level perspectives. As an exploration of the subject and a tool to target technical research, the widespread adoption of generative AI technologies is examined as a product of societal trends and policy decisions through which the evolution of AI is being shaped. Along with this general investigation, a more technical examination of performance and feasibility of computing hardware for large language models is analyzed through a comparison of architectures such as Apple Silicon and NVIDIA CUDA. This research aids in predicting the trajectory of GenAI as a technological system and supports development of efficient subsystems to improve the environmental sustainability of AI. Through these projects, a commitment is shown to ensuring that the future of AI is developed in a way that is powerful, efficient, and mindful of human and ecological needs.

School of Engineering and Applied Science

Bachelor of Science in Computer Science

Technical Advisor: Felix Lin

STS Advisor: Bryn Seabrook

Technical Team Members: Tao Groves