[Journal]
“ROSETTA: A Resource and Energy-Efficient Inference Processor for Recurrent Neural Networks Based on Programmable Data Formats and Fine Activation Pruning”
Author: *Jiho Kim and Tae-Hwan Kim
To be published in IEEE Trans. on Emerging Topics in Computing (TETC), 2022
“Low-Latency Bearing Fault Diagnosis Based on Convolutional LSTM Model” (in Korean)
Author: *Jiho Kim, Jihoon Shin, and Taehwan Kim
Published in Jrnl. of the Korean Institute of Electronics and Information Engineers (IEIE), 2022
“AERO: A 1.28 MOP/s/LUT Reconfigurable Inference Processor for Recurrent Neural Networks in a Resource-Limited FPGA”
Author: Jinwon Kim, *Jiho Kim, and Taehwan Kim
Published in Electronics, 2021
[Conference]
“Understanding the Performance and Estimating the Cost of LLM Fine-Tuning,”
Author: Yuchen Xia, *Jiho Kim, Yuhan Chen, Haojie Ye, Souvik Kundu, Cong (Callie) Hao, and Nishil Talati
Published in IEEE International Symposium on Workload Characterization (IISWC), 2024
“An Integrated Verification System for an RNN Inference Processor” (in Korean)
Author: Seongjae Hong, *Jiho Kim, and Taehwan Kim
Published in Summer Conf. of the Korean Institute of Electronics and Information Engineers (Conf. of IEIE), 2022
“A Reconfigurable Inference Processor for Recurrent Neural Networks Based on Programmable Data Format in a Resource-Limited FPGA”
Author: *Jiho Kim, Kwoanyoung Park, and Taehwan Kim
Published in Asia and South Pacific Design Automation Conf. (ASP-DAC), 2022
“Low-Latency Bearing Fault Diagnosis Based on Convolutional Long Short-Term Memory” (in Korean)
Author: *Jiho Kim, Jihoon Shin, and Taehwan Kim
Published in Summer Conf. of the Korean Institute of Electronics and Information Engineers (Conf. of IEIE), 2021
[Patent]
“Dynamic Pruning Apparatus and Method for Instruction-set Based Neural Network Inference Processor”
Inventor: *Jiho Kim, and Taehwan Kim
Type: Korea Patent
Application number: -
Filing date: 2022-06-30
Legal status: Pending
“Apparatus and Method for Hardware Implementation of Activation Function in Recurrent Neural Network”
Inventor: *Jiho Kim, and Taehwan Kim
Type: Korea Patent
Application number: -
Filing date: 2021-04-29
Legal status: Pending
“Apparatus and Method for Improving Inference Speed of Neural Network Model”
Inventor: *Jiho Kim, and Taehwan Kim
Type: Korea Patent
Application number: 10-2020-0052268
Filing date: 2020-04-29
Legal status: Issued