Yashasvi Makin has more than 10 years of experience working in software, leading engineering teams. He specializes in software engineering, optimizing GPUs for machine learning recommendation models, making large-scale infrastructure more efficient, and building scalable AI systems.

Title : Green AI at Scale: Hardware–Software Co-Design for Low-Carbon Training and Compiler Optimization -  Abstract: How do we make frontier AI both faster and greener? This talk shares production-tested techniques like kernel fusion, mixed precision, sharded optimizers, memory-and data-pipeline tuning, and carbon-aware orchestration across NVIDIA/AMD/other hardware that raise throughput per watt. Drawing on real deployments, it offers a playbook for cutting energy use and cost while enabling cross-disciplinary research.