| Author | Khosla, Nitin |
| Call Number | AIT Thesis no. CS-92-18 |
| Subject(s) | Integrated circuits--Very large scale integration
|
| Note | A thesis submitted in partial fulfillment of the requirements for the degree of Master of
Engineering, School of Engineering and Technology |
| Publisher | Asian Institute of Technology |
| Abstract | Cell placement is an important step in the physical design of VLSI circuits. It is a complex
optimization problem which has been proved NP-hard. The present day algorithms are inherently
sequential and rely on rigorous algorithmic techniques.
Self-Organi~ation scheme is an attractive neural net technique which can be applied to
solve such a ยท combinatorial VLSI cell placement problem in an effective manner. The selforganization scheme is an unsupervised self-learning scheme where the weights of synapses
between the output nodes and input nodes are adjusted to select topological relationships between
the input end and the output end. The main feature in this method is to find the most sensitive
output node (called as the Winning Node) to a given input pattern and to choose a neighborhood
around the winning node. The adjustments of the weights inside this neighborhood selected,
influences the outcome iteratively to the other input patterns. Although it is known that the choice
of neighborhood effects the rate of convergence, the relationship between convergence and
neighborhood is not known in quantitative terms. At the same time, the gain values during the
process do play an important role in the convergence of the alg01ithm. Problem size also
influences greatly the rate of convergence of the algorithm. In this thesis we have examined
extensively the choice of neighborhoods & their effects in the placement of cells, variation of the
gain values during the process of cells placement, suitable range of gain values and the effect of
problem size on the rate of convergence of cells placement process. The results are generalizable
to other domains. |
| Year | 1992 |
| Type | Thesis |
| School | School of Engineering and Technology (SET) |
| Department | Department of Information and Communications Technologies (DICT) |
| Academic Program/FoS | Computer Science (CS) |
| Chairperson(s) | Sadananda, Ramakoti |
| Examination Committee(s) | Huynh Ngoc Phien ;Kanchana Kanchanasut; |
| Scholarship Donor(s) | Canadian International Development Agency; |
| Degree | Thesis (M.Eng.) - Asian Institute of Technology, 1992 |