Due to their greater economy, gigahertz computers will eventually replace megahertz computers. Circuits and system organizations have been actively explored for the last 3 years, and various devices and circuits have been proposed. This paper attempts to integrate and give perspective, from a system viewpoint, to isolated disclosures, by classifying and contrasting various developments with emphasis on the more promising approaches. The first gigahertz computer will probably be made of tunnel-diode baseband circuitry. For the long range, thin-film cryotrons, with a cost advantage of 100 to 1 over tunnel diodes, show the greatest promise.