Advanced Computing System Laboratory (Kondo Lab)
Department of Information and Computer Science,
Keio University
About this site
This site introduces recent research activities of Kondo Lab in the department of Information and Computer Science, Keio University.
Our lab is tightly cooperated with Amano Lab in the same department.
Research summary
We are researching on domain-specific architectures which are optimized for specific application domains such as graph processing and robot intelligence, and computer systems that combine the benefits of classical computing and new computing paradigms such as AI/ML, neuromorphic computing, and quantum computing. Specific research themes are listed below.
Computer architecture
Graph processing framework for smart society
Cryogenic computing
Next Generation Supercomputer Systems
Computer architecture
The computer architecture research seeks for better organization of processing core such as CPU, GPU, and accelerators. Our interests in this topic are developing accelerator engines and its algorithms for AI, graph processing, and robot intelligence as well as evaluation of them on real LSIs and FPGAs.
CPS framework for smart society
Cyber Physical Systems (CPS) are one of the most important aspects in realizing smart society. AI and Graph processing are powerful tools in CPS for modeling and optimizing complex problems. We are conducting research on a graph processing framework, edge-cloud cooperation, and on-device learning technique for anomaly detection.
Cryogenic computing
Novel computing principles and systems based on the superconducting devices in cryogenic environment are becoming crucial technologies. We are researching new computer systems with the single flux quantum technology and its application to control logic for quantum computers.
Next-generation supercomputer systems
In designing next generation supercomputer systems, their power consumption becomes a first class design constraint. This research focuses on power-efficient HPC architecture, power-aware job scheduling, and efficient algorithm to solve scheduling problems.
Access
Department of Information and Computer Science,
Faculty of Science and Technology (Yagami-Campus),
Keio University
3-14-1 Hiyoshi, Kohoku-ku, Yokohama, Kanagawa 223-8522, Japan.