KAIST EPIC Lab

Welcome to EPIC LAB at KAIST EE

Embedded Processor & Intelligent Computing (EPIC) Lab is a research group in the School of Electrical Engineering at Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Korea. With a passionate team, EPIC Lab develops optimized hardware/software building blocks for modern application-specific embedded processors targeting a wide range of intelligent computing workloads.

Our research spans cross-disciplinary topics including algorithms, architectures, and system-on-chip (SoC) designs. Current focus areas include advanced deep learning processors for compressed machine learning models, high-performance baseband processors for next-generation communication systems, and energy-efficient IoT platforms for human-centric applications.

Latest News

[2025/11] Our paper entitled "IterQuant: Iterative quantization framework for mixed-precision LLM compression" has been accepted for presentation at IEEE/ACM Design, Automation and Test in Europe (DATE 2026).

[2025/11] Our paper entitled "A 3.3 Gb/s/mm^2 area-efficient non-binary LDPC decoder using column-layered processing" has been accepted for publication in the IEEE Transactions on Circuits and Systems I: Regular Papers. 

[2025/10] Our research on the URLLC polar decoder has been selected as the Chip Design Contest Best Poster Award at the 2025 International SoC Design Conference (ISOCC 2025).

[2025/10] Our research on the FPGA-based ISP accelerator has been selected as the Best Live Demo Award at the 2025 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS 2025).

[2025/09] Our paper entitled "NMIX: NoC-aware mixed precision quantization for energy-efficient ML accelerator" has been accepted for publication in the IEEE Transactions on Circuits and Systems II: Express Briefs.

[2025/09] Jaehee has received the Student Travel Grants by IEEE CASS for his paper "Hybrid ordered statistics decoding of short-length BCH codes for URLLC systems: Theoretical analysis and decoder implementation" at IEEE ISICAS 2025.