General Competition (April 1992)

Polymeric Switches for Optical Interconnects

Sponsor: IBM Corporation, Almaden Research Center

• Project Performance Period: 8/1/1992 - 7/31/1995
• Total project (est.): $3,565,254.00
• Requested ATP funds: $1,838,788.00

Because of their ability to carry huge amounts of data with relative immunity from external interference, optical fibers have become an essential technology in electronic communications, for example to link large computer systems. In the future, optical fibers or waveguides are expected to link smaller computers in distributed processing systems, individual circuit boards within a computer, individual chips on a circuit board, and perhaps even individual elements within a single chip. If so, the industry will require far more compact and efficient electro-optical switches to route data on the optical fibers. The current generation of switches convert the data from an optical to an electrical signal, do the necessary switching, and then convert the data back to an optical signal, a process that involves expensive components and significantly limits the speed of the system. IBM, working with the University of Colorado, proposes to develop a switch technology based on recently developed non-linear optical polymers that would work without the necessity of converting the signal. Such optical polymer switches could be formed with photolithography on semiconductor substrates in a process compatible with standard integrated circuit manufacturing. The planned program will develop the necessary polymer materials, electro-optical switches using the new polymers, and an engineering computer-aided design (CAD) tool to aid efficient product design. The total switch industry is quite large -- expected to reach $30 billion within the decade -- and optical switches will capture an increasingly large share of that market.

For project information:

Michael Ross, (408) 927-1283

ATP Project Manager

Michael Schen, (301) 975-6741
michael.schen@nist.gov