I. Summaries A. Spatial Computing for Supporting Geographic Science
Download 11.26 Kb.
|
- Navigate this page:
- I.Summaries
- B.Adiabatic Quantum Computation
- C.From Atoms to Material Properties: Basic Physics in the Development of New Nano-materials
- II.Question Discussion
- References
Report of three presentations on graduate seminar Fenye Bao Abstract—This report concluded three presentations on the graduate seminar of Department of CS, Virginia Tech (NVC). Index Terms—summary, seminar, Virginia Tech I.SummariesA.Spatial Computing for Supporting Geographic ScienceThis presentation was given by Professor Chaowei Yang on September 4, 2009. In geographic information science, the analysis and simulations are computing intensive, which is hardly to be supported by single computer. During this talk, Prof. Yang introduced the spatial computing and its objective. Spatial computing refers to the computing that utilizes the spatial principles and distributed computers for enabling the advancements of geographic sciences. The hardware infrastructure they built for spatial computing is a distributed computing environment among several universities and institutions in the United States. He also presented three research examples in geographic sciences which are supported by spatial computing. The first application example is the geographic data and service discovery and access. The second example is utilizing HPC (High Performance Computing) in the distributed environment to support the computing in geographic applications. The third example is the geospatial data visualization supported by spatial computing. He demoed the projects of data and service discovery and geospatial data visualization. The infrastructure for spatial computing is mainly supported by high performance computing and grid computing technologies which heavily depend on the network routing and transactions and the granularity of the grid. Prof. Yang also demonstrated that how the performance of the system is affected by the network routing and the size of the grid. Best on their evaluation to the system, the optimal grid size can be found. Some of their work can be found at: [1-2]. B.Adiabatic Quantum ComputationThis presentation was given by Professor Vicky Choi on September 25, 2009. Quantum Computation can be utilized to solve NP-hard problems by designing quantum algorithms. Adiabatic Quantum Computation (AQC) is a quantum model which has been proven to be equivalent to conventional quantum computation. Prof. Vicky Choi presented the maximum independent set problem (MIS, which is a maximum problem), the Ising Problem (which is a minimum problem), and the correspondence of solving the Max problem by Min problem. She briefly introduced the Quantum Computation. A classical bit is 0 or 1, which a Qubit (quantum bit) is a superposition of the quantum |x> = C0|0> + C1|1>, |C0|2 + |C1|2 = 1. The measurement of |x> is 0 with the probability of |C0|2 and 1 with the probability of |C1|2. The classical computation is discrete, which the Adiabatic Quantum Computation is continuous. The Adiabatic Quantum Computation utilizes Hamiltonian which is a 2n by 2n complex matrix. The MIS problem and factoring problem can be solved in polynomial time with AQC algorithm, causing the widely used encryption methods, like RAS etc. to be broken. There are also fundamental issues of AQC to be solved in analysis and design of AQC algorithms, How to design a “good” initial Hamiltonian? How to design a best encoded final Hamiltonian? How to design the evolution path? Her work on this topic can be found in paper [3]. She also has a US patent[4] in this field.
Materials are mostly designed by nature. The physics and material science development bring the concept of “materials by design”. Prof. Farkas firstly introduced the multiscale physics in the dimensions of time and length. Their research focuses on the scale between Quantum mechanics (QC) and Molecular Dynamic (MD). The basic concept is the interatomic potential. The interatomic potential will increase when the distance between atoms is either too far or too close. Their research includes the visualization and simulation of material properties with the clustering computing environment in Virginia Tech. They simulated and visualized the material cracks in 3-dimensions with different structures. Their experiments illustrate that the path of material crack follows some patterns formed by the defects in nano-scale inside the material and the material becomes harder when increasing the stress, but not hard again if the stress beyond a certain threshold. Finally, she concluded that the current computing power allows the simulation of material properties and the simulation can reach smallness as tested in the real materials. Prof. Farkas also pointed that the computer clusters at Virginia Tech is very powerful and convenient to use. Today, there are many software libraries available for simulation in this area. It’s easier than years ago to conduct the experiments even for new researchers. II.Question DiscussionQuestion: How far is the applicable quantum computation from us? Answer: There was a quantum computing talked with title “Quantum Computing – how far have we come, and where will we end?” We may still have image of the appearance of the first real quantum computer which depends on physics, computer science, electronic engineering, etc. ReferencesDownload 11.26 Kb. Share with your friends:
|