Browsing by Author "Chen, Cheng-Kai"

Now showing 1 - 4 of 4

A distributed simulation-based computational intelligence algorithm for nanoscale semiconductor device inverse problem
(2006)
Li, Yiming
;
Chen, Cheng-Kai
;
電信工程研究所
;
Institute of Communications Engineering
In this paper, a distributed simulation-based computational intelligence algorithm for inverse problem of nanoscale semiconductor device is presented. This approach features a simulation-based optimization strategy, and mainly integrates the semiconductor process simulation, semiconductor device simulation, evolutionary strategy, and empirical knowledge on a distributed computing environment. For a set of given target current-voltage (I-V) curves of metal-oxide-semi conductor field effect transistors (MOSFETs) devices, the developed prototype executes evolutionary tasks to solve an inverse doping profile problem, and therefore optimize fabrication recipes. In the evolutionary loop, the established management server allocates the jobs of process simulation and device simulation on a PC-based Linux cluster with message passing interface (MPI) libraries. Good benchmark results including the speed-up, the load balancing, and the parallel efficiency are presented. Computed results, compared with the realistic measured data of 65 nm n-type MOSFET, show the accuracy and robustness of the method.
A simulation-based evolutionary technique for inverse doping profile problem of sub-65 nm CMOS devices
(2006-12-01)
Li, Yiming
;
Chen, Cheng-Kai
;
電信工程研究所
;
Institute of Communications Engineering
In this paper, we utilize an evolutionary technique for inverse doping profile problems of the 65 nm complementary metal oxide semiconductor (CMOS) devices. The approach mainly bases upon the process simulation, device simulation, evolutionary strategy, and empirical knowledge. For a set of given measured I-V curves of the 65 nm CMOS, a developed prototype performs the optimization task to automatically calibrate and inversely search out, for example the doping recipe and device physical model parameters. The simulation-optimization-coupled methodology is complicated theoretically, but our preliminary results imply that it may benefit the development of fabrication technology and can be used for the performance diagnosis, in particular, for sub-65 nm devices.
A unified optimization framework for microelectronics industry
(2006)
Li, Yiming
;
Chen, Cheng-Kai
;
Cho, Yen-Yu
;
電信工程研究所
;
Institute of Communications Engineering
In this paper, an object-oriented unified optimization framework (UOF) for general problem optimization is proposed. Based on evolutionary algorithms, numerical deterministic methods, and C++ objective design, the UOF possesses real-world applications for various optimization problems. The UOF is with interface between defining a general problem and generic solver; hence the two areas are bridged. The UOF's components ate categorized into problem and solver parts, and they work independently. Therefore, the reusable high level code allows the adaptation to new problem and solver quickly. Practical applications in microelectronics industry confirm the ftinctionality of the developed unified optimization framework.
Visi - A VTK- and QT-Based Open-Source Project for Scientific Data Visualization
(2009)
Li, Yiming
;
Chen, Cheng-Kai
;
電信工程研究所
;
Institute of Communications Engineering
In this paper, we present an open-source project, Visi for high-dimensional engineering and scientific data visualization. Visi is with state-of-the-art interactive user interface and graphics kernels based upon Qt (a cross-platform GUI toolkit) and VTK (an object-oriented visualization library). For an initialization of Visi, a preliminary window will be activated by Qt, and the kernel of VTK is simultaneously embedded into the window, where the graphics resources are allocated. Representation of visualization is through an interactive interface so that the data will be rendered according to user's preference. The developed framework possesses high flexibility and extensibility for advanced functions (e.g., object combination, etc) and further applications. Application of Visi to data visualization in various fields, such as protein structure in bioinformatics, 3D semiconductor transistor, and interconnect of very-large scale integration (VLSI) layout is also illustrated to show the performance of Visi. The developed open-source project is available in our project website on the intemet [1].