World-largest class materials simulations using JAIST supercomputers
MAEZONO Laboratory
Professor:MAEZONO Ryo
E-mail:
[Research areas]
Materials Informatics & Condensed matter theory using High Performance Computing
[Keywords]
Many-body electron theory/Quantum Chemistry/Condensed Matter Physics/Materials Informatics/Diffusion Monte Carlo/Ab initio methods
Skills and background we are looking for in prospective students
Maths taught in under-grad. common educations, English language skill required to survey journals, Adaptability to research interests in fundamental field of materials science, Sociability required for international research activities. Intellectual curiosity without sticking to one's background.
What you can expect to learn in this laboratory
We are working on 'ab initio' approaches excluding arbitrary modeling as possible to keep objectivity. Research activities requires not only logical/theoretical capabilities but also such toughness to develop several possible discussions for a scenario on your data which comes entirely from fundamental equations without any prior modeling for a picture. Scrupulous surveying over preceding studies is the required task which enables you to come up with how to develop your discussions, what points to be tackled, and how your contributions are positioned in the whole efforts made by researches so far. As we take it as the primary educational policy, such capabilities are those required for any leaders in the projects done at any organizations such as companies.
【Job category of graduates】
ICT sections at heavy industries and electric power companies, major industries of materials/chemicals, full researchers in national research laboratories overseas, post-docs in abroad etc.
Research outline
【Materials Informatics Science】
Frontiers of Information Science has enabled the massive data handling with reasonable speed, and has made qualitative changes in materials science recently. The simulation handles vast amount of couplings from at most 100 kinds of realistic atoms on the earth with quantum mechanical calculations to predict materials properties those are useful for our life. Such activities are about to be promoted intensively by several countries as their scientific strategies. The world of quantum mechanics is out of our intuitive expectations and hence the power of computer simulation is essentially in demand.
【Quantum many-body theory and Massive Parallel Simulations】
We have worked on the statistical simulations involving random number samplings combined with quantum many-body theory (Diffusion Monte Carlo method). The framework provides the most reliable estimation of electronic properties of materials. Our simulations require massive computational resources, but inherently in harmony with parallel processing. JAIST's super computing facilities enable us to carry out such simulations. JAIST is a representative center of HPC in Japan, possessing four kinds of super computers.
【Interdisciplinary Challenge】
There is a huge demand for researchers working on physics, chemistry, and materials science, who are good not only at mathematics but at 'computics' as another tool of analyzing phenomena. It is a great challenge for talented younger students to commit such a novel, exciting interdisciplinary field as a 'Simulation Scientist', experienced by rich computational facilities and fine course work provided by JAIST.
Key publications
See our group webpage for the publications.
We've published in such journals as JACS(IF=13.858)/
Angewande(IF=11.709)/Phys.Rev.Lett.(7.435)/
JCTC(5.39)/Sci.Rep.(5.228)/Inorg.Chem.(4.857)/
Phys.Rev.B(3.767)/J.Chem.Phys.(3.164)/
J.Comput.Chem.(3.84)/Appl.Phys.Lett.(3.794)
Equipment
- Supercomputers (CRAY XC40/SGI Altix etc.)
- World fragship class computers (K/Japan, Theta/US etc.)
- Lab-own handmade parallel computers and file storages.
- Customized Macintosh PC's with maximum efficiency.
Teaching policy
We're putting importance on such capabilities required for those who can lead the society as a respectable 'degree holder'. These include the abilities [1/to plan projects with sufficient assessment on feasibility and sustainability, not just by impulsive motivations], [2/to make decisions on the comparison over the possible options based on the scientific criticism, not sticking an opinion], [3/to manage over several missions running parallelwise], [4/to enjoy pursuing well brushed up skills for document managements, digesting, and presentations with rich/appealing vocabularies.
[Website] URL:https://www.jaist.ac.jp/is/labs/maezono-lab/homepage2019/index.html