摘要 |
<p>One embodiment of the present invention is a method for constructing defect-and-failure-tolerant demultiplexers (figures 14 an 16). This method is applicable to nanoscale, microscal, or larger-scale demultiplexer circuits,. Demultiplexer circuits can be viewed as a set of AND gates (figures 9A-B), each including a reversibly switchable interconnection between a number of address lines (910-912 and 920-922), or address-line-derived signal lines, and an output signal line (914 and 924). Each reversibly switchable interconnection includes one ot more reversibly switchable elements (906-908 and 916-918). In certain demultiplexer embodiments, NMOS (102) and/or PMOS transistors (206) are employed as reversibly switchable elements. In the method that representd one embodiment of the present invention, two or more serially connected transistors (410, 412, and 411, 413; 1502) are employed in each reversibly switchable interconnection, so that short defects in up to one less then the number of serially interconnected transistors does not lead to failure of the reversibly switchable interconnection. In addition, error-control-encoding techniques are used to introduce additional address-line-derived signal lines (1602, 1604) and additional switchable interconnections (1610) so that the demultiplexer may function even when a number of individual, switchable interconnections are open-defective.</p> |
申请人 |
HEWLETT-PACKARD DEVELOPMENT COMPANY, L. P.;ROBINETT, WARREN;KUEKES, PHILIP, J.;WILLIAMS, STANLEY, R. |
发明人 |
ROBINETT, WARREN;KUEKES, PHILIP, J.;WILLIAMS, STANLEY, R. |