Research
Our research area includes thermal-aware/low-power system design for AI, robust memory systems, next-generation memory architecture, architecture-level thermal modeling, etc. These research topics are being actively studied in the industry as well as academia.
Research Interests
Various Topics in Computer Architecture, Memory Systems, etc.
Edge AI System Design/Optimization
Robust Memory Systems
Next-generation Memory (FeRAM, MRAM, ReRAM, PRAM, etc.)
Application-specific Performance Optimization
Thermal-aware System Optimization
edge ai system design/optimization
Pruning/Quantization/Knowledge Distillation
Processing-in-Memory (PIM), Computation-in-Memory (CIM)
Peripheral logic design for Inference/Training Optimization
LLM Inference/Training Optimization at the Edge
Robust memory systems
Error-tolerable Encoding for AI
Memory Error Profiling for Various Workloads
Error Correction Code (e.g., BCH, Reed-solomon, etc.)
Memory Controller/On-die ECC for HBM
Next-generation Memory Architecture
Ferroelectric Memory
STT-MRAM (Spin-Transfer Torque Magnetoresistive RAM)
ReRAM (Resistive RAM)
Application-specific Computing Systems
Domain-specific Architecture
Processing-in-Memory (PIM)
Application-aware Computing Platform
GPU-based SW Acceleration and Parallelization
Thermal-aware System Optimization
Thermal-aware Microarchitecture Design
Thermal Modeling Methodology
Thermal Sensor Allocation