Tuesday, 13 June 2006

Nanotechnologies offer the potential to improve memory density by several orders of magnitude but only at the expense of introducing stochastic variations in chip parameters that are much larger than are tolerable with lithographic-based assembly. In this talk we describe methods of dealing with such variation. In particular, we examine the crossbar, an old idea, that is expected to play a prominent role in these new developments. Data will be stored at the crosspoints

defined by orthogonal nanowires (NWs) in the two dimensions of the crossbar. One NW in each dimension will be activated by applying fields to lithographic-scale wires (LWs). For a nanometer-scale crossbar to use area efficiently, it is not possible to activate one NW in each dimension by attaching one LW to each NW. Instead it will be necessary to control each NW using a small number of LWs that are common to all NWs. Each of the proposed methods for controlling NWs with LWs involves stochastic assembly. In this talk we describe four such methods and compare their performance based on statistical analysis.

John E. Savage earned the Sc.B, Sc.M and Ph.D. degrees in Electrical Engineering from MIT in 1962, 1962 and 1965, respectively. From 1965 to 1967 he was a Member of Technical Staff at Bell Laboratories in Holmdel, NJ. He has been a member of the Brown University Faculty since 1967 and since 1979 has been a member of the Department of Computer Science, of which he is a founder; he was Chair of that department from 1985 until 1991. John is a Guggenheim Fellow, a Life Fellow of the IEEE, and a Fellow of ACM and AAAS. He was a member of the MIT Corporation Visiting Committee for the Department of Electrical Engineering and Computer Science from 1991 to 2002. At Brown University he has served as Vice Chair, Chair and Past Chair of the Faculty from 2000-2003, Chair of the Task Force on Faculty Governance from 2002-2003 and as a member or chair of many other university-level committees. He is the author of more than 70 research publications. John is the author of two books on theoretical computer science The Complexity of Computing published in 1976 by JohnWiley and Sons and Models of Computation published in 1998 by Addison-Wesley. He also is co-editor with T. Knight, of the book Advanced Research in VLSI and Parallel Systems, published by the MIT Press in 1992.

The meeting will be held at the Lincoln Lab Auditorium at 4:00 PM. Refreshments will be served at 3:30 PM.  Registration is in the main lobby.  Foreign national visitors to Lincoln Lab require visit requests.  You may also pre-register by e-mail to Roslyn Wesley, rwesley@ll.mit.edu; please indicate your citizenship. You will not receive a confirmation of your pre-registration, however, your badge will be ready for you when you register.   Please use the Wood Street Gate.  For directions go to http://www.ll.mit.edu For other information, contact Ed Altshuler, Chairman at (781)377-4662 or edward.altshuler@hanscom.af.mil