Author: webmaster

Hello! I’m Koda, a second-year doctoral student.

At the Doctoral Student Presentation Session held on July 17, I presented my research entitled “代理モデルを用いたX線タイコグラフィの高速化に関する研究” and received the Best Presentation Award!

This award recognizes presentations that communicate the presenter’s research clearly and engagingly to an audience from a wide range of academic backgrounds.
X-ray ptychography may be unfamiliar to many people, so I am delighted that I was able to convey what makes this field interesting and that my presentation was recognized.

Encouraged by this award, I will continue working hard to achieve even better research outcomes!

Hi, I’m Haoran Ding, a second-year Master’s student.

Recently, Chang, Pan and I have just finished the final defense for my graduation thesis, thanks to the support of the professors and all the lab members. I could not have completed this thesis without everyone’s help.

Also, I have submitted my first draft to the Administration Office, and I will submit the polished version in the near future. I hope it turns out well!

Hi, This is Specially Appointed Assistant Professor Tao!

From July 9 to 10, 2026, we attended the 18th JHPCN Symposium at Tokyo Conference Center Shinagawa. JHPCN, the Joint Usage/Research Center for Interdisciplinary Large-scale Information Infrastructures, brings together information infrastructure centers at eight universities in Japan to promote interdisciplinary research across computational science, computer science, and data science. The symposium provided a forum for sharing the latest research results and plans from JHPCN projects.

The symposium featured a wide range of topics, including the operation of high-performance computing (HPC) systems, job scheduling and storage technologies, artificial intelligence and machine learning, weather and climate prediction, simulations in materials and life sciences, computing for disaster prevention, and quantum-HPC integration. In an invited talk titled “Quantum-Centric Supercomputing: Current Status and Future Prospects,” Hiroshi Horii of IBM Quantum Japan discussed the current state and future of next-generation computing platforms that flexibly integrate quantum processing units (QPUs), graphics processing units (GPUs), and central processing units (CPUs). From our laboratory, Prof. Takizawa presented his work on an urgent computing infrastructure that leverages supercomputers across multiple sites. The symposium reaffirmed the importance of applying advanced computing resources and system software to real-world challenges in interdisciplinary computational research. I hope to apply the insights gained to my future research and to user support at the Cyberscience Center.

Hello! I’m Watanabe, a first-year master’s student (M1).

We recently held a welcome party for the new B3 students who have joined the Takizawa Laboratory!

Although the weather was less than ideal with light rain, we enjoyed having a barbecue together. It was also a great opportunity to get to know the new B3 members, and we all had a wonderful time.

To all of our new B3 students, welcome to the Takizawa Laboratory! We hope you’ll enjoy both your research and our laboratory activities, and we look forward to spending a fulfilling and enjoyable time together.

Hi, this is Specially Appointed Assistant Professor Tao!

From May 18 to 20, 2026, I attended the 37th NEC User Group Meeting, NUG Society Meeting XXXVII 2026, held at the Information Technology Center of the University of Cologne (ITCC) in Cologne, Germany. The NEC User Group (NUG) is an international meeting where users, operators, and developers of NEC supercomputers and high performance computing (HPC) systems share the latest technologies, operational experience, and research results.

From our laboratory, Prof. Takizawa and I gave talks. Prof. Takizawa spoke on AI-assisted programming in his talk titled “Towards AI-assisted Programming at the Tohoku University Cyberscience Center.” He introduced HPC user support activities at the Tohoku University Cyberscience Center and discussed the potential of large language models (LLMs) for code understanding, performance optimization, semantic equivalence verification, automated debugging, and performance bottleneck analysis. I gave a talk titled “A Runtime Prediction Model for Capacity Planning in Shared-Backend LLM Multi-Agent Social Simulation,” introducing AI-agent social simulation for tsunami evacuation. Based on the supercomputer AOBA and ExpressHPC, an urgent computing framework that connects multiple supercomputers, I discussed how to predict the runtime of large-scale LLM multi-agent simulations and use it for capacity planning in city-scale evacuation simulation.

The meeting also covered a wide range of topics, including NEC’s HPC and Research Information Infrastructure (HPC/RII), next-generation vector systems, RISC-V and vector extensions, GPU resource utilization, integration of HPC and AI services at RWTH Aachen University, Slurm cluster operation, storage research at the Center for Computational Sciences, University of Tsukuba, and high-performance fluid dynamics algorithms at the University of Cologne. Through this meeting, I was reminded that the future development of HPC depends on the combined progress of hardware, software, user support, and AI applications. I hope to apply the insights gained from this meeting to my future research and to user support at the Cyberscience Center.