The Experts below are selected from a list of 37002 Experts worldwide ranked by ideXlab platform
Helmut G. Katzgraber - One of the best experts on this subject based on the ideXlab platform.
-
physics inspired optimization for quadratic unconstrained problems using a digital annealer
Frontiers of Physics in China, 2019Co-Authors: Maliheh Aramon, Gili Rosenberg, Elisabetta Valiante, Toshiyuki Miyazawa, Hirotaka Tamura, Helmut G. KatzgraberAbstract:The Fujitsu Digital Annealer is designed to solve fully connected quadratic unconstrained binary optimization (QUBO) problems. It is implemented on Application-specific CMOS hardware and currently solves problems of up to 1024 variables. The Digital Annealer's algorithm is currently based on simulated annealing; however, it differs from it in its utilization of an efficient parallel-trial scheme and a dynamic escape mechanism. In addition, the Digital Annealer exploits the massive parallelization that Custom Application-specific CMOS hardware allows. We compare the performance of the Digital Annealer to simulated annealing and parallel tempering with isoenergetic cluster moves on two-dimensional and fully connected spin-glass problems with bimodal and Gaussian couplings. These represent the respective limits of sparse versus dense problems, as well as high-degeneracy versus low-degeneracy problems. Our results show that the Digital Annealer currently exhibits a time-to-solution speedup of roughly two orders of magnitude for fully connected spin-glass problems with bimodal or Gaussian couplings, over the single-core implementations of simulated annealing and parallel tempering Monte Carlo used in this study. The Digital Annealer does not appear to exhibit a speedup for sparse two-dimensional spin-glass problems, which we explain on theoretical grounds. We also benchmarked an early implementation of the Parallel Tempering Digital Annealer. Our results suggest an improved scaling over the other algorithms for fully connected problems of average difficulty with bimodal disorder. The next generation of the Digital Annealer is expected to be able to solve fully connected problems up to 8192 variables in size. This would enable the study of fundamental physics problems and industrial Applications that were previously inaccessible using standard computing hardware or special-purpose quantum annealing machines.
-
Data_Sheet_1_Physics-Inspired Optimization for Quadratic Unconstrained Problems Using a Digital Annealer.pdf
2019Co-Authors: Maliheh Aramon, Gili Rosenberg, Elisabetta Valiante, Toshiyuki Miyazawa, Hirotaka Tamura, Helmut G. KatzgraberAbstract:The Fujitsu Digital Annealer is designed to solve fully connected quadratic unconstrained binary optimization (QUBO) problems. It is implemented on Application-specific CMOS hardware and currently solves problems of up to 1,024 variables. The Digital Annealer's algorithm is currently based on simulated annealing; however, it differs from it in its utilization of an efficient parallel-trial scheme and a dynamic escape mechanism. In addition, the Digital Annealer exploits the massive parallelization that Custom Application-specific CMOS hardware allows. We compare the performance of the Digital Annealer to simulated annealing and parallel tempering with isoenergetic cluster moves on two-dimensional and fully connected spin-glass problems with bimodal and Gaussian couplings. These represent the respective limits of sparse vs. dense problems, as well as high-degeneracy vs. low-degeneracy problems. Our results show that the Digital Annealer currently exhibits a time-to-solution speedup of roughly two orders of magnitude for fully connected spin-glass problems with bimodal or Gaussian couplings, over the single-core implementations of simulated annealing and parallel tempering Monte Carlo used in this study. The Digital Annealer does not appear to exhibit a speedup for sparse two-dimensional spin-glass problems, which we explain on theoretical grounds. We also benchmarked an early implementation of the Parallel Tempering Digital Annealer. Our results suggest an improved scaling over the other algorithms for fully connected problems of average difficulty with bimodal disorder. The next generation of the Digital Annealer is expected to be able to solve fully connected problems up to 8,192 variables in size. This would enable the study of fundamental physics problems and industrial Applications that were previously inaccessible using standard computing hardware or special-purpose quantum annealing machines.
-
Physics-Inspired Optimization for Quadratic Unconstrained Problems Using a Digital Annealer
Frontiers Media S.A., 2019Co-Authors: Maliheh Aramon, Gili Rosenberg, Elisabetta Valiante, Toshiyuki Miyazawa, Hirotaka Tamura, Helmut G. KatzgraberAbstract:The Fujitsu Digital Annealer is designed to solve fully connected quadratic unconstrained binary optimization (QUBO) problems. It is implemented on Application-specific CMOS hardware and currently solves problems of up to 1,024 variables. The Digital Annealer's algorithm is currently based on simulated annealing; however, it differs from it in its utilization of an efficient parallel-trial scheme and a dynamic escape mechanism. In addition, the Digital Annealer exploits the massive parallelization that Custom Application-specific CMOS hardware allows. We compare the performance of the Digital Annealer to simulated annealing and parallel tempering with isoenergetic cluster moves on two-dimensional and fully connected spin-glass problems with bimodal and Gaussian couplings. These represent the respective limits of sparse vs. dense problems, as well as high-degeneracy vs. low-degeneracy problems. Our results show that the Digital Annealer currently exhibits a time-to-solution speedup of roughly two orders of magnitude for fully connected spin-glass problems with bimodal or Gaussian couplings, over the single-core implementations of simulated annealing and parallel tempering Monte Carlo used in this study. The Digital Annealer does not appear to exhibit a speedup for sparse two-dimensional spin-glass problems, which we explain on theoretical grounds. We also benchmarked an early implementation of the Parallel Tempering Digital Annealer. Our results suggest an improved scaling over the other algorithms for fully connected problems of average difficulty with bimodal disorder. The next generation of the Digital Annealer is expected to be able to solve fully connected problems up to 8,192 variables in size. This would enable the study of fundamental physics problems and industrial Applications that were previously inaccessible using standard computing hardware or special-purpose quantum annealing machines
Partha Roy Chaudhuri - One of the best experts on this subject based on the ideXlab platform.
-
near elliptic core triangular lattice and square lattice pcfs a comparison of birefringence cut off and gvd characteristics towards fiber device Application
arXiv: Optics, 2014Co-Authors: Partha Sona Maji, Partha Roy ChaudhuriAbstract:In this work, detailed numerical analysis of the near-elliptic core index-guiding triangular-lattice and square-lattice photonic crystal fiber (PCFs) are reported for birefringence, single mode, cut-off behavior, group velocity dispersion and effective area properties. For the same relative values of d/P, triangular-lattice PCFs show higher birefringence whereas the square-lattice PCFs show a wider range of single-mode operation. Square-lattice PCF was found to be endlessly single-mode for higher air-filling fraction (d/P). Smaller lengths of triangular-lattice PCF are required for dispersion compensation whereas PCFs with square-lattice with nearer relative dispersion slope (RDS) can better compensate the broadband dispersion. Square-lattice PCFs show ZDW red-shifted, making it preferable for mid-IR supercontinuum generation (SCG) with highly non-linear chalcogenide material. Square-lattice PCFs show higher dispersion slope that leads to compression of the broadband, thus accumulating more power in the pulse. On the other hand, triangular-lattice PCF with flat dispersion profile can generate broader SCG. Square-lattice PCF with low Group Velocity Dispersion (GVD) at the anomalous dispersion corresponds to higher dispersion length and higher degree of solitonic interaction. The effective area of square-lattice PCF is always greater than its triangular-lattice counterpart making it better suited for high power Applications. Smaller length of symmetric-core PCF for dispersion compensation, while broadband dispersion compensation can be better performed with asymmetric-core PCF. Mid-Infrared SCG can be better performed with asymmetric-core PCF with compressed and high power pulse, while wider range of SCG can be performed with symmetric core PCF. Thus, this study will be extremely useful for realizing fiber towards a Custom Application around these characteristics.
-
near elliptic core triangular lattice and square lattice pcfs a comparison of birefringence cut off and gvd characteristics towards fiber device Application
Journal of The Optical Society of Korea, 2014Co-Authors: Partha Sona Maji, Partha Roy ChaudhuriAbstract:In this work, we report detailed numerical analysis of the near-elliptic core index-guiding triangular-lattice and square-lattice photonic crystal fiber (PCFs); where we numerically characterize the birefringence, single mode, cut-off behavior and group velocity dispersion and effective area properties. By varying geometry and examining the modal field profile we find that for the same relative values of $d/{\Lambda}$ , triangular-lattice PCFs show higher birefringence whereas the square-lattice PCFs show a wider range of single-mode operation. Square-lattice PCF was found to be endlessly single-mode for higher air-filling fraction ( $d/{\Lambda}$ ). Dispersion comparison between the two structures reveal that we need smaller lengths of triangular-lattice PCF for dispersion compensation whereas PCFs with square-lattice with nearer relative dispersion slope (RDS) can better compensate the broadband dispersion. Square-lattice PCFs show zero dispersion wavelength (ZDW) red-shifted, making it preferable for mid-IR supercontinuum generation (SCG) with highly non-linear chalcogenide material. Square-lattice PCFs show higher dispersion slope that leads to compression of the broadband, thus accumulating more power in the pulse. On the other hand, triangular-lattice PCF with flat dispersion profile can generate broader SCG. Square-lattice PCF with low Group Velocity Dispersion (GVD) at the anomalous dispersion corresponds to higher dispersion length ( $L_D$ ) and higher degree of solitonic interaction. The effective area of square-lattice PCF is always greater than its triangular-lattice counterpart making it better suited for high power Applications. We have also performed a comparison of the dispersion properties of between the symmetric-core and asymmetric-core triangular-lattice PCF. While we need smaller length of symmetric-core PCF for dispersion compensation, broadband dispersion compensation can be performed with asymmetric-core PCF. Mid-Infrared (IR) SCG can be better performed with asymmetric core PCF with compressed and high power pulse, while wider range of SCG can be performed with symmetric core PCF. Thus, this study will be extremely useful for designing/realizing fiber towards a Custom Application around these characteristics.
Chih-hong Wong - One of the best experts on this subject based on the ideXlab platform.
-
Leveraging Cloud Platform for Custom Application Development
2011 IEEE International Conference on Services Computing, 2011Co-Authors: Nianjun Zhou, Da Peng An, Liang-jie Zhang, Chih-hong WongAbstract:Compared with packaged Application, Custom Application developments (CAD) experience the frustration of higher project overhead and less certainty. The typical time spent on building the infrastructure for a CAD project is, on average, several weeks. Project uncertainty comes from unique Customer requirements and lack of standardized methods and toolsets to follow. Therefore, a CAD project is more difficult to achieve cost reduction and asset reuse. In this paper, we present a cloud platform to alleviate this problem through an integration of a) standard methods, b) standardized toolsets aligned with those methods, c) project management environments with pre-defined work breakdown structure (WBS) aligned with those methods and toolsets, and d) infrastructure support from the cloud technology. We believe that such a cloud platform will become a fundamental approach for large enterprises to develop CAD or other solutions for their clients.
Partha Sona Maji - One of the best experts on this subject based on the ideXlab platform.
-
near elliptic core triangular lattice and square lattice pcfs a comparison of birefringence cut off and gvd characteristics towards fiber device Application
arXiv: Optics, 2014Co-Authors: Partha Sona Maji, Partha Roy ChaudhuriAbstract:In this work, detailed numerical analysis of the near-elliptic core index-guiding triangular-lattice and square-lattice photonic crystal fiber (PCFs) are reported for birefringence, single mode, cut-off behavior, group velocity dispersion and effective area properties. For the same relative values of d/P, triangular-lattice PCFs show higher birefringence whereas the square-lattice PCFs show a wider range of single-mode operation. Square-lattice PCF was found to be endlessly single-mode for higher air-filling fraction (d/P). Smaller lengths of triangular-lattice PCF are required for dispersion compensation whereas PCFs with square-lattice with nearer relative dispersion slope (RDS) can better compensate the broadband dispersion. Square-lattice PCFs show ZDW red-shifted, making it preferable for mid-IR supercontinuum generation (SCG) with highly non-linear chalcogenide material. Square-lattice PCFs show higher dispersion slope that leads to compression of the broadband, thus accumulating more power in the pulse. On the other hand, triangular-lattice PCF with flat dispersion profile can generate broader SCG. Square-lattice PCF with low Group Velocity Dispersion (GVD) at the anomalous dispersion corresponds to higher dispersion length and higher degree of solitonic interaction. The effective area of square-lattice PCF is always greater than its triangular-lattice counterpart making it better suited for high power Applications. Smaller length of symmetric-core PCF for dispersion compensation, while broadband dispersion compensation can be better performed with asymmetric-core PCF. Mid-Infrared SCG can be better performed with asymmetric-core PCF with compressed and high power pulse, while wider range of SCG can be performed with symmetric core PCF. Thus, this study will be extremely useful for realizing fiber towards a Custom Application around these characteristics.
-
near elliptic core triangular lattice and square lattice pcfs a comparison of birefringence cut off and gvd characteristics towards fiber device Application
Journal of The Optical Society of Korea, 2014Co-Authors: Partha Sona Maji, Partha Roy ChaudhuriAbstract:In this work, we report detailed numerical analysis of the near-elliptic core index-guiding triangular-lattice and square-lattice photonic crystal fiber (PCFs); where we numerically characterize the birefringence, single mode, cut-off behavior and group velocity dispersion and effective area properties. By varying geometry and examining the modal field profile we find that for the same relative values of $d/{\Lambda}$ , triangular-lattice PCFs show higher birefringence whereas the square-lattice PCFs show a wider range of single-mode operation. Square-lattice PCF was found to be endlessly single-mode for higher air-filling fraction ( $d/{\Lambda}$ ). Dispersion comparison between the two structures reveal that we need smaller lengths of triangular-lattice PCF for dispersion compensation whereas PCFs with square-lattice with nearer relative dispersion slope (RDS) can better compensate the broadband dispersion. Square-lattice PCFs show zero dispersion wavelength (ZDW) red-shifted, making it preferable for mid-IR supercontinuum generation (SCG) with highly non-linear chalcogenide material. Square-lattice PCFs show higher dispersion slope that leads to compression of the broadband, thus accumulating more power in the pulse. On the other hand, triangular-lattice PCF with flat dispersion profile can generate broader SCG. Square-lattice PCF with low Group Velocity Dispersion (GVD) at the anomalous dispersion corresponds to higher dispersion length ( $L_D$ ) and higher degree of solitonic interaction. The effective area of square-lattice PCF is always greater than its triangular-lattice counterpart making it better suited for high power Applications. We have also performed a comparison of the dispersion properties of between the symmetric-core and asymmetric-core triangular-lattice PCF. While we need smaller length of symmetric-core PCF for dispersion compensation, broadband dispersion compensation can be performed with asymmetric-core PCF. Mid-Infrared (IR) SCG can be better performed with asymmetric core PCF with compressed and high power pulse, while wider range of SCG can be performed with symmetric core PCF. Thus, this study will be extremely useful for designing/realizing fiber towards a Custom Application around these characteristics.
Nianjun Zhou - One of the best experts on this subject based on the ideXlab platform.
-
Leveraging Cloud Platform for Custom Application Development
2011 IEEE International Conference on Services Computing, 2011Co-Authors: Nianjun Zhou, Da Peng An, Liang-jie Zhang, Chih-hong WongAbstract:Compared with packaged Application, Custom Application developments (CAD) experience the frustration of higher project overhead and less certainty. The typical time spent on building the infrastructure for a CAD project is, on average, several weeks. Project uncertainty comes from unique Customer requirements and lack of standardized methods and toolsets to follow. Therefore, a CAD project is more difficult to achieve cost reduction and asset reuse. In this paper, we present a cloud platform to alleviate this problem through an integration of a) standard methods, b) standardized toolsets aligned with those methods, c) project management environments with pre-defined work breakdown structure (WBS) aligned with those methods and toolsets, and d) infrastructure support from the cloud technology. We believe that such a cloud platform will become a fundamental approach for large enterprises to develop CAD or other solutions for their clients.
