Is Finer Better? The Limits of Microscaling Formats in Large Language ModelsAndrea FasoliMonodeep Karet al.2026ICLR 2026Conference paper
Spyre: An inference-optimized scalable AI accelerator for enterprise workloadsMatt CohenMonodeep Karet al.2026ISSCC 2026Conference paper
Hardware Accelerator Design for AI: Enabling Generative ModelsLeland Chang2025VLSI Technology and Circuits 2025Short course
Architecture and Design Approaches to ML Hardware Acceleration: Performance Compute EnvironmentLeland Chang2024ISSCC 2024Short course