Embedded Files
Hardware-Aware AI Algorithm Compression
Hardware-Aware AI Algorithm Compression
Goal: Lightweight AI Algorithm Design Optimized for Current AI Hardware Constraints
We develop versatile quantization algorithms applicable across diverse Generative AI models to ensure broad hardware compatibility. Additionally, we design adaptive algorithms that dynamically prioritize data based on its importance, significantly reducing computational overhead while maintaining high accuracy.
[HPCA 2026] Sangjin Kim, Yuseon Choi, Byeongcheol Kim, Jungjun Oh, Hoi-jun Yoo
“GyRot: Leveraging Hidden Synergy between Rotation and Fine-grained Group Quantization for Low-bit LLM Inference,” in IEEE International Symposium on High-Performance Computer Architecture[JETCAS 2025] Sangjin Kim, Yuseon Choi, Jungjun Oh, Byeongcheol Kim, Hoi-Jun Yoo
“LightRot: A Light-weighted Rotation Scheme and Architecture for Accurate Low-bit Large Language Model Inference,” in IEEE Journal on Emerging and Selected Topics in Circuits and Systems[ISSCC 2026] Sangjin Kim, Jungjun Oh, Byeongcheol Kim, Yuseon Choi, Gwangtae Park, Hoi-jun Yoo
“Revolver: Low-Bit GenAI Accelerator for Distilled-Model and CoT with Phase-Aware-Quantization and Rotation-Based Integer-Scaled Group Quantization,” in IEEE International Solid-State Circuits Conference
Page updated
Google Sites
Report abuse