RIKEN Center for Advanced Intelligence Project AI Computing Team

Team Director: Shinya Takamaeda (D.Eng.)

Japanese Page

Research Summary

As the demand for computational power and associated costs for deep learning continues to rise, energy-efficient and high-performance computing technologies are essential for further evolution and practical application of machine learning. Our team explores the design and realization method of advanced machine learning systems by working across multiple layers, including circuits, devices, architectures, and algorithms. Specifically, we investigate energy-efficient computer architecture for machine learning based on novel devices and computing principles, device-aware machine learning algorithms, and system architectures for large language model (LLM) inference serving that achieve low latency and high bandwidth with minimal energy consumption.

Main Research Fields

Informatics

Related Research Fields

Engineering
Computer Architecture
Distributed Processing
Machine Learning

Keywords

AI Chip
Algorithm/ Hardware Co-design
Domain-Specific Architecture
Compute in Memory
Computing with Emerging Devices

Selected Publications

Papers with an asterisk(*) are based on research conducted outside of RIKEN.

1. *Daichi Tokuda and Shinya Takamaeda-Yamazaki
"DF-BETA: An FPGA-based Memory Locality Aware Decision Forest Accelerator via Bit-Level Early Termination"
ACM Transactions on Reconfigurable Technology and Systems, Vol.18, Issue.2, pp.1-26, 2025.
2. *Honoka Anada, Ryu Sefutsu, Masayuki Usui, Tatsuya Kaneko, and Shinya Takamaeda-Yamazaki
"PRIOT: Pruning-Based Integer-Only Transfer Learning for Embedded Systems"
IEEE Embedded Systems Letters, 2024.
3. *Yuki Hirayama and Shinya Takamaeda-Yamazaki
"Scalable Moment Propagation and Analysis of Variational Distributions for Practical Bayesian Deep Learning"
IEEE Transactions on Neural Networks and Learning Systems, Vol.36, Issue.3, pp.4614-4624, 2025.
4. *Tatsuya Kubo, Masayuki Usui, Tomoya Nagatani, Daichi Tokuda, Lei Qu, Ting Cao, and Shinya Takamaeda-Yamazaki
"Bulk Bitwise Accumulation in Commercial DRAM"
NeurIPS 2024 Workshop Machine Learning with new Compute Paradigms (MLNCP 2024), 2024.
5. *Wenlun Zhang, Shimpei Ando, Yung-Chin Chen, Satomi Miyagi, Shinya Takamaeda-Yamazaki, and Kentaro Yoshioka
"PACiM: A Sparsity-Centric Hybrid Compute-in-Memory Architecture via Probabilistic Approximation"
2024 ACM/IEEE International Conference on Computer-Aided Design (ICCAD 2024), 2024.
6. *Yung-Chin Chen, Shimpei Ando, Daichi Fujiki, Shinya Takamaeda-Yamazaki, and Kentaro Yoshioka
"OSA-HCIM: On-The-Fly Saliency-Aware Hybrid SRAM CIM with Dynamic Precision Configuration"
29th Asia and South Pacific Design Automation Conference (ASP-DAC 2024), 2024.
7. *Masayoshi Tsutsui, Tatsuya Kaneko, and Shinya Takamaeda-Yamazaki
"Poison Egg: Scrambling Federated Learning with Delayed Backdoor Attack"
3rd International Conference on Ubiquitous Security 2023 (UbiSec-2023), 2023.
8. *Kasho Yamamoto, Kota Ando, Normann Mertig, Takashi Takemoto, Masanao Yamaoka, Hiroshi Teramoto, Akira Sakai, Shinya Takamaeda-Yamazaki, and Masato Motomura
"STATICA: A 512-Spin 0.25M-Weight Full-Digital Annealing Processor with a Near-Memory All-Spin-Updates-at-Once Architecture for Combinatorial Optimization with Complete Spin-Spin Interactions"
2020 International Solid-State Circuits Conference (ISSCC 2020), 2020.
9. *Kodai Ueyoshi, Kota Ando, Kazutoshi Hirose, Shinya Takamaeda-Yamazaki, Junichiro Kadomoto, Tomoki Miyata, Mototsugu Hamada, Tadahiro Kuroda, and Masato Motomura
"QUEST: A 7.49-TOPS Multi-Purpose Log-Quantized DNN Inference Engine Stacked on 96MB 3D SRAM using Inductive-Coupling Technology in 40nm CMOS"
2018 International Solid-State Circuits Conference (ISSCC 2018), 2018.
10. *Shinya Takamaeda-Yamazaki
"Pyverilog: A Python-based Hardware Design Processing Toolkit for Verilog HDL"
11th International Symposium on Applied Reconfigurable Computing (ARC 2015), pp.451-460, 2015.

Lab Members

Principal investigator

Shinya Takamaeda: Team Director

Contact Information

Nihonbashi 1-chome Mitsui Building, 15th floor,
1-4-1 Nihonbashi, Chuo-ku, Tokyo
103-0027, Japan
Email: shinya.takamaeda@riken.jp