WE PROPOSE TO STUDY AND DEVELOP SPACE QUALIFIED ELECTRO-OPTICAL DEVICES THAT ARE CAPABLE OF ULTRA HIGH SPEED SIGNAL PROCESSING AND LOGIC OPERATION AND ARE COMPATIBLE WITH THE PULSE-POWER AND REPETITION RATES AVAILABLE WITH PICOSECOND PULSES FROM A

WE PROPOSE TO STUDY AND DEVELOP SPACE QUALIFIED ELECTRO-OPTICAL DEVICES THAT ARE CAPABLE OF ULTRA HIGH SPEED SIGNAL PROCESSING AND LOGIC OPERATION AND ARE COMPATIBLE WITH THE PULSE-POWER AND REPETITION RATES AVAILABLE WITH PICOSECOND PULSES FROM A MODE-LOCKED SEMICONDUCTOR DIODE LASER, WHICH WE HAVE DEVELOPED OVER THE PAST FEW YEARS. SUCH OPTICAL LOGIC DEVICES, BASED ON PICOSECOND PROBLEMS RELATED TO THE COUPLING OF HIGH-SPEED SIGNALS TO AND FROM THE GAAS CHIP (AND THE RELATED PROBLEMS OF PARASITICS AND LOADING), BUT ALSO PROVIDE AN UNAMBIGUOUS DETERMINATION OF THE GATE DELAY (IN CONTRAST WITH THE CONVENTIONAL RING OSCILLATOR AND GIGAHERTZ RATE "DIVIDED-BYTWO" TECHNIQUES). NUMEROUS VARIATIONS OF THESE OPTICAL TECHNIQUES WILL BE IMPLEMENTED IN OUR LABORATORIES FOR THE DIRECT MEASUREMENT OF ON-CHIP PROPAGATION DELAYS AND WAVEFORMS IN GIGAHERTZ LOGIC CIRCUITS WITH UNPRECEDENTED (\ 10 PS) RESOLUTION AND ACCURACY. THE POTENTIAL FOR SUCH CONCEPTS FOR NOVEL DESIGNS OF COMPUTER ARCHITECTURE WILL BE EXPLORED. AS CONTINUED IMPROVEMENTS IN ELECTRONIC SPEED AND POWER BEGIN TO SLOW, OPTICS WILL ASSUME A ROLE OF INCREASING IMPORTANCE IN OUR QUEST TO MAINTAIN COMPUTATIONAL SUPERIORITY IN THE U.S. OVER OUR RIVALS IN BOTH THE ECONOMIC AND NATIONAL SECURITY REALMS.