Commercial Applications
High-End Computing
Low temperature superconductor technology can take computing speed far beyond the theoretical limits of silicon while simultaneously effecting major reductions in both size and power requirements. While highly challenging, the path forward is clear and success will enable the most complex and interactive tasks to be performed that cannot be addressed in any other way.
Computers have increased exponentially in performance each year for more than 30 years, sustained by exponential increases in processor gate count and chip speed. In recent years, Moore's Law advances have slowed due to unsustainable chip heating required to simultaneously support more gates and faster clocks. The consequence is a limit in processor speeds to about 3 to 4 GHz, but ever-increasing gate count allows more processors on a single chip.
This new direction maintains performance growth for games, home and business computing, and even supercomputing for problems which can be made parallel or otherwise broken up. It is proving inadequate for problems that cannot be broken into small computational pieces. Unfortunately, many important national security problems cannot be easily broken up and separately managed. These problems require use of the sustained, full capability of a supercomputer.
Superconducting electronics lead to both low gate power and ultra-high speed, with a theoretical limit estimated to be potentially 100 times faster than silicon. Already, superconductor-based Rapid Single Flux Quantum (RSFQ) circuitry has easily broken the 20 GHz speed barrier and speeds of 100 GHz are predicted within the next five years. Extremely fast switching time, operation at very low voltages, and very little power consumed or dissipated could provide a breakthrough for supercomputing for critical national needs.
In 2004, the Federal Plan for High-End Computing recognized the need to develop new technologies to meet emerging security threats and identified superconductor electronics as the fastest digital processors of any electronic technology. Subsequently, an extensive detailed technical and needs analysis by the National Security Agency's Office of Corporate Assessments, the Superconducting Technology Assessment, concluded that superconductive technology is an excellent candidate for computing at petaflops (one million billion operations per second) and beyond. The report points out that while there is considerable development and demonstration to be done, no new research breakthroughs are required. Superconductor electronics can help to close the growing gap between peak performance and sustained system performance in high end computers and a major US Government push to mature this technology is needed.